xvideos: Normalize the URL or it will fail with some inputs.

For instance, if we give it <www.xvideos.com/video1384059>, we would
end up passing that to urllib2, which would complain about an unknown
URL type.

diff --git a/youtube-dl b/youtube-dl
index 450da8e..f8e9095 100755 (executable)
--- a/youtube-dl
+++ b/youtube-dl
@@ -3413,7 +3413,7 @@ class XVideosIE(InfoExtractor):
 
                self.report_webpage(video_id)
 
-               request = urllib2.Request(url)
+               request = urllib2.Request(r'http://www.xvideos.com/video' + video_id)
                try:
                        webpage = urllib2.urlopen(request).read()
                except (urllib2.URLError, httplib.HTTPException, socket.error), err: