Sort of nitpicky, but I think you can have Flask automatically decode the data using the correct encoding scheme (as set in the HTTP header) with r.get_data(as_text=True). It is just a test and I doubt it'll get run on a system that's not using UTF-8 by default, but I think technically it's not safe to assume the data is UTF-8 encoded.

Same comment as above

Also here

If you want to continue importing these to provide a nice package interface (I do like having the __init__.py define the public interface) you can mark the specific lines to be ignored by the linter:

from .base import db  # noqa: F401
from .waivers import Waiver  # noqa: F401

will explicitly ignore imports that are unused, or you can tell flake8 to ignore any error with a plain # noqa comment.

http://flake8.pycqa.org/en/latest/user/ignoring-errors.html#in-line-ignoring-errors for reference

Generally +1, but I recommend using the get_data API.

Cool, I like it.

Right.

Yes! Looks good to me too. :+1: