Working on my latest project, todoster, forced me to learn all about input/output mocking using python's built-in unittest library.

Mocking Output to stdout or stderr

The way I've been using the mock output or error stream is by getting its contents as a string. For this, there is a very simple method:

mock_err.getValue()
mock_out.getValue()

With this string you can do assertions as normal. However, since mock_err and mock_output aren't actually strings but streams you need to be very careful when trying to use the same stream a second time. The streams don't get cleared automatically so you need to either re-declare the mocks or make sure they're manually cleared out before every re-use.

If you want to have one mock for one test case (class) just make sure you're resetting mock_out and mock_err after every test like so:

import patch from unittest.mock

class TestClass
    mock_out = patch('sys.stdout', new_callable=StringIO)
    mock_err = patch('sys.stderr', new_callable=io.StringIO)

    def tearDown(self):
        self.mock_out.truncate(0)
        self.mock_out.seek(0)
        self.mock_err.truncate(0)
        self.mock_err.seek(0)

    # ...

On the other hand, you can mock on a test by test basis by using decorators like so:

from unittest.mock import patch

class TestClass(unittest.TestCase):
    #...

    @patch('sys.stderr', new_callable=io.StringIO)
    @patch('sys.stdout', new_callable=io.StringIO)
    test_small_functionality(self, mock_out, mock_err)
        # test something

Now you don't have to worry about clearing out the mocks after every test but it does come with having to add a new decorator for every test method that needs the mock.

Mocking stdin

If you want to mock stdin, this is the simplest solution I could find. It assumes that the code under test calls input only once though.

@patch('builtins.input', side_effect=['the input you want to test'])
def test_delete_project_say_no(self, _):
    # test something