Unit Test

Template

import unittest
from unittest.mock import patch

class MyTestCase(unittest.TestCase):

  def test_my_method(self):
    self.assertEqual(1, 1)

  @patch('my.module.program.func')
  def test_my_method_with_path(self, mock_func):
    mock_func.return_value = 3

if __name__ == '__main__':
    unittest.main()

Patching

We should always patch the variables in the scope of the program we’re testing, even if itself imported them from somewhere else. Example:

my/module/program.py:

from lib import func

It imported func from lib.

test.py:

...
@patch('my.module.program.func')
def test_my_method(self, mock_func):
  pass

We patch func from my.module.program, not lib.

Return Different Values

If we assign a list to side_effect it returns the i-th value each time the function is called.

my/module/program.py:

def f():
  return 1

def g():
  print(f(), f(), f())

from my.module.program import g
...

@patch('my.module.program.f')
def test_my_method(self, mock_func):
  mock_func.side_effect = [1, 2, 3]
  g() # prints 1 2 3

Mock Implementation

Assign a function to side_effect:

def f(i):
  return i

def g(i):
  print(f(i))

from my.module.program import g
...

@patch('my.module.program.f')
def test_my_method(self, mock_func):
  def mock_func_impl(i):
    return i*2
  mock_func.side_effect = mock_func_impl
  g(3) # prints 6

Mock Throwing Exception

Assign a function to side_effect:

from my.module.program import g
...

@patch('my.module.program.f')
def test_my_method(self, mock_func):
  mock_func.side_effect = RuntimeError()
  g(3) # throws