HTTP and Swift transports should retry connection on failure

[alistairking]

Currently if the client reads data very slowly, the server may time out the connection, and we'll get an error like:

Traceback (most recent call last):
  File "<stdin>", line 2, in <module>
  File "build/bdist.linux-x86_64/egg/wandio/file.py", line 22, in next
  File "build/bdist.linux-x86_64/egg/wandio/compressed.py", line 68, in next
  File "build/bdist.linux-x86_64/egg/wandio/compressed.py", line 81, in readline
  File "build/bdist.linux-x86_64/egg/wandio/compressed.py", line 31, in _refill
  File "build/bdist.linux-x86_64/egg/wandio/file.py", line 25, in read
  File "/usr/local/lib/python2.7/dist-packages/swiftclient/client.py", line 311, in read
    buf = self.resp.read(length)
  File "/usr/local/lib/python2.7/dist-packages/swiftclient/client.py", line 463, in releasing_read
    chunk = self.resp.raw.read(*args, **kwargs)
  File "/usr/local/lib/python2.7/dist-packages/urllib3/response.py", line 401, in read
    raise IncompleteRead(self._fp_bytes_read, self.length_remaining)
  File "/usr/lib/python2.7/contextlib.py", line 35, in __exit__
    self.gen.throw(type, value, traceback)
  File "/usr/local/lib/python2.7/dist-packages/urllib3/response.py", line 320, in _error_catcher
    raise ProtocolError('Connection broken: %r' % e, e)
urllib3.exceptions.ProtocolError: ("Connection broken: error(104, 'Connection reset by peer')", error(104, 'Connection reset by peer'))

We should catch this and re-open the connection and then seek to the offset of the data previously consumed.

[alistairking]

This could also happen if the server drops the connection for some other reason (e.g., some failure on its end), so we could try resuming a few times and then raise an exception so that the user can decide what to do.