new prints ascii decode error

[Brant Knudson]

`reno new` prints an error before it creates the file:

$ .tox/venv/bin/reno new keystone-change-3
ERROR: 'ascii' codec can't decode byte 0xc6 in position 0: ordinal not in range(128)
Created new notes file in releasenotes/notes/keystone-change-3-af64d3a8dafe14e0.yaml

[Brant Knudson (blk-u)]

proposed fix is https://review.openstack.org/#/c/243816/

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/243816
Committed: https://git.openstack.org/cgit/openstack/reno/commit/?id=0feb86ecbb0b77a83da42a5a5add959c7834ffff
Submitter: Jenkins
Branch: master

commit 0feb86ecbb0b77a83da42a5a5add959c7834ffff
Author: Brant Knudson <email address hidden>
Date: Tue Nov 10 14:23:21 2015 -0600

    Fix unicode handling in get_random_string

    get_random_string would often fail in decoding the result of
    os.urandom. It would fail because the result of urandom is a bunch
    of random bytes and not valid text (i.e., not a valid utf-8
    representation).

    Turns out that it's unnecessary to decode the result of os.urandom.
    It returns six.binary_type whether on py2 or py3, and
    binascii.hexlify() takes a six.binary_type as the argument.

    binascii.hexlify() returns six.binary_type and we want
    get_random_string() to return six.text_type, so convert.

    Also, fixed the docstrings to use """.

    Closes-Bug: 1514981
    Change-Id: I60c31f2c47614beb3ed3c143d7815fdc15f5a0e0