Unicode Characters in Tags Cause Problems

[cshoe]

Seen in datadogpy 0.9.0 and 0.11.0 using Python 2.7.6 and 2.7.11.

It doesn't seem possible to use unicode characters in tags. Running the following causes an UnicodeDecodeError to be raised from the use of str here:

# -*- coding: utf-8 -*-

from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

from datadog import statsd

statsd.increment('tagged_counter', 1, tags=['person:Jóhn', 'foo:bar'])

If the tags are encoded before they are sent to datadogpy, the error is raised in 0.11.0 a couple lines before when the payload is being constructed.

# -*- coding: utf-8 -*-

from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

from datadog import statsd

statsd.increment('tagged_counter', 1, tags=['person:Jóhn'.encode('utf-8'), 'foo:bar'])

[yannmh]

Thanks for reporting the issue @cshoe.

I am adding it to our 0.12.0 milestone, so it gets fixed in the next release.