This is caused by the fact that thepiratebay's search responds with a redirect to an URI containing whatever was searched, but doesn't properly encode the URI, so you get something like this: HTTP/1.1 302 Found Soup-Debug-Timestamp: 1237071670 Soup-Debug: SoupMessage 7 (0xa159f10) Set-Cookie: searchTitle=0; expires=Sun, 14-Mar-2010 23:01:19 GMT; path=/ Location: /search/foo bar/0/99/0 Content-Type: text/html Content-Length: 0 Date: Sat, 14 Mar 2009 23:01:19 GMT Server: lighttpd libsoup currently uses the URI with no processing to make sure it's valid, and gets a 400 Bad Request.
Created attachment 28631 [details] proposed fix
(In reply to comment #1) > Created an attachment (id=28631) [review] > proposed fix > + SoupURI* soup_uri = soup_uri_new(newURL.string().utf8().data()); + soup_message_set_uri(msg, soup_uri); + You are leaking the new URI, need to call soup_uri_free on it.
Created attachment 28634 [details] proposed fix Yep, Xan's write =).
Comment on attachment 28634 [details] proposed fix r=me
Landed as r41719.
(In reply to comment #3) > Created an attachment (id=28634) [review] > proposed fix > > Yep, Xan's write =). > right is what I meant, of course... /me hits his head against the wall =)
Created attachment 33171 [details] webkit-bug-24602-drop-workaround.patch Drop workaround required for <=libsoup-2.26.2
Created attachment 33172 [details] webkit-bug-24602-drop-workaround.patch
Comment on attachment 33172 [details] webkit-bug-24602-drop-workaround.patch looks good r=me
Comment on attachment 33172 [details] webkit-bug-24602-drop-workaround.patch Landed as r46178.