Unicode handling

[brutasse]

In various places py-trello .encode('utf-8') some strings returned by the API.

What's the rationale behind this? Trello data is returned in JSON, with string keys and attributes in unicode strings (u'foo' in Python 2, 'foo' in Python 3).

Keeping unicode would make sense: in Python 2 you can compare bytestrings to unicode strings implicitly (u'foo' == 'foo'), but in Python 3 it'd make more sense to compare e.g. board.name to unicode strings instead of bytestrings. In Python 3 our code needs to re-decode data for comparison or string formatting. We have things like:

if board.name == b'Board name'

or:

message = '{}: {}'.format(board.name.decode(), card.name.decode())

My suggested fix would be to remove all occurrences of .encode('utf-8') here, to have a proper unicode sandwich behavior. Any objections? I'm happy to submit a PR.

[brutasse]

Refs #66 #118

[sfriesel]

this is the first commit that introduced the encoding: c925101 It was triggered by a wrong implementation of repr in the previous commit 2914118

[sarumont]

No objections to proper UTF-8 fixes. I merged the previous PR, as I didn't have the time to properly review it. :) PRs are welcome!

[sfriesel]

choose your poison :)

[sarumont]

Thanks for the work, guys! I merged #138, and I bumped the version to 0.6.0. Anyone who is depending on the old, broken behavior can still use 0.5.1. #139 appears to have been a subset of #138, so I won't worry about that unless I missed something @sfriesel.

[brutasse]

Thanks for the merge, @sarumont!