.. _exercise_mailroom_mocking:

################
Mocking Mailroom
################

Goal
====

You should now have a nice mailroom program, complete with an object-oriented structure and bundled up into a nice python package.

And a complete set of unit tests for the logic code.

But your command line user interface code is likely not tested. It's hard to auto-test user-interaction.

Your mission is to get mailroom fully tested.

Start with your object oriented mailroom in a proper python package.

Coverage
--------

First, run coverage on your current tests. I like ``pytest-cov``:

.. code-block:: bash

    $ pip install pytest-cov

    $ pytest --cov=mailroom

Or, to get the nifty html output:

.. code-block:: bash

    $ pytest --cov=mailroom --cov-report html

That will result in a pile of html in an ``htmlcov`` directory. Point your browser to the ``index.html`` file in there, and click away.

Once you've run coverage, add tests to get it up close to 100% on your logic code, if it's not there already.

Fixtures
--------

Fixtures are a really good way to make your tests cleaner and more independent. With mailroom, you should have a couple fixtures that set up a donor database with some data in it, and maybe one or two for ``Donor`` objects.

Clean up your tests with fixtures, and keep the coverage up!

One possible use for a fixture is providing a file to write to for the code that writes files. Or maybe a directory to put the files in, and then it can clean up the dir in teardown.

Mocking input
-------------

Once you have 100% coverage for the logic code, it's time to test the UI.

You should be able to use mocking to mock the ``input`` function, and then actually test your user interface code, too.

Can you get 100% test coverage?