Like it says on the tin -- adds a bunch of tests for the List functions and refactors/fixes some of the old stuff that didn't get picked up before with py3 compatibility. There are a couple things of note:

• Some methods, namely GetListsList don't require a user_id or screen_name parameter -- if neither is passed, then the data will be for the authenticated user.
• New methods/functions: type checking can be done with enf_type, which takes the name of the field, the type the field should be, and the value to convert and raises the appropriate errors if it cannot be done. So instead of:

    try:
        parameters['count'] = int(count)
    except ValueError:
        raise TwitterError({'message': '"count" must be an integer'})

it's just:

    parameters['count'] = enf_type('count', int, count)

The error raised is largely similar, it tells you the name of the field that was incorrect and the expected type. This isn't standard across the API since I wanted to keep this PR to just the List functions, but that's a todo I guess if it seems rational.

• Refactored out ID'ing a specific list into its own static method _IDList() which takes the list-y kwargs and figures out how the user is trying to identify a list (i.e., by id or slug + (user_id/screen_name).

I think that's about it, but I've seen the word "list" about a million times these two days and I'm sure I missed something.

I also alphabetized the migration guide.