unittest_mock

Import

from unittest.mock import Mock, patch

Usage

Production Code

def child_function(input: str):
    return input

def parent_function(input: str):
    return child_function(input)

Test Code

from unittest.mock import Mock, patch
from full.module.path import child_function

class MyTest(unittest.TestCase):
    @patch("full.module.path.child_function")
    def test_parent_function(self, mock_child_function):
        # When `child_function()` gets called, even indirectly,
        # the mock object will be called instead,
        # and it will return "my_string".
        mock_child_function.return_value = "my_string"
        result = parent_function()
        mock_child_function.assert_called_once()
        self.assert(result, "my_string")
        
        mock_child_function.reset_mock() # reset return value, etc.
        # Everytime `child_function()` is called,
        # even indirectly, it will raise this exception.
        mock_child_function.side_effect = ValueError("Mocked exception")
        with self.assertRaises(ValueError):
            parent_function()
        mock_child_function.assert_called_once()

Patched Function Path

For the full function namespace path in @patch(), you need to use the namespace of the file that contains the function you're testing, not the file where the patched function is defined.

Example:

• file_1.py contains function def add(a, b)

• file_2.py contains function def add_and_print(), which calls function add()

• You're writing a unit test for function add_and_print() in a separate test.py file

• When you patch add(), you would use the namespace of file_2.py like:

  • @patch("path.to.file_2.add")

• Even though the function is defined at path.to.file_1, when it gets imported into file_2.py, it takes on the namespace of file_2.py