Don't crash when attempting to decode an invalid URL

[f0rk]

2014-10-29 16:15:54,681 could not decode url: 'http://www.test.com/\xc0.\xc0./\xc0.\xc0./\xc0.\xc0./\xc0.\xc0./windows/win.ini'
Traceback (most recent call last):
 File "/usr/lib/python2.7/dist-packages/pyramid_exclog/__init__.py", line 111, in exclog_tween
   return handler(request)
 File "/usr/lib/python2.7/dist-packages/pyramid_tm/__init__.py", line 82, in tm_tween
   reraise(*exc_info)
 File "/usr/lib/python2.7/dist-packages/pyramid_tm/__init__.py", line 62, in tm_tween
   t.note(request.path_info)
 File "/usr/lib/python2.7/dist-packages/webob/descriptors.py", line 68, in fget
   return req.encget(key, encattr=encattr)
 File "/usr/lib/python2.7/dist-packages/webob/request.py", line 178, in encget
   return val.decode(encoding)
 File "/usr/lib/python2.7/encodings/utf_8.py", line 16, in decode
   return codecs.utf_8_decode(input, errors, True)
UnicodeDecodeError: 'utf8' codec can't decode byte 0xc0 in position 1: invalid start byte

Discovered with the help of Nessus.

[mmerickel]

Good find. Can you please submit a test for this to maintain 100% coverage?

[f0rk]

How? The use of fake requests and dummy transactions makes that a bit difficult. Obviously I can raise an exception in the appropriate place but there's so much mocked out it's not like it reflects reality.

[mmerickel]

The test is to see how pyramid_tm handles an un-decodable path_info. Simply make a dummy request that raises on the path_info property to provide a justification for that code path existing.

[f0rk]

Tests added

[mmerickel]

Looks great, thank you very much.

[tseaver]

Hmm, looks like we don't have Travis running on PRs. This change broke tests on Py3k: I've pushed a change which fixes it