How to Mock External Services in pytest Using unittest.mock

When writing unit tests with pytest , you often need to mock external services or functions. The standard library module unittest.mock provides powerful mocking capabilities that integrate smoothly with pytest .

1. Install dependencies

Make sure pytest is installed (the unittest.mock module is included in the Python standard library).

pip install pytest

2. Create the module to be tested

Assume a file my_module.py with a function fetch_data that calls an external API.

# my_module.py
import requests

def fetch_data(url):
    """Fetch data from the given URL.
    Args:
        url (str): API endpoint URL.
    Returns:
        dict: JSON data returned by the API.
    Raises:
        requests.exceptions.HTTPError: If the HTTP request fails.
    """
    response = requests.get(url)
    response.raise_for_status()  # Raise if the request failed
    return response.json()

3. Create the test file

In test_my_module.py , import pytest , MagicMock , patch , and the function to test.

# test_my_module.py
import pytest
from unittest.mock import MagicMock, patch
from my_module import fetch_data

3.1 Define a fixture to set up the mock object

# Define a fixture to set up the mock object
@pytest.fixture
def mock_requests_get():
    """Use a pytest fixture to mock
requests.get
."""
    with patch('requests.get') as mock_get:
        yield mock_get

3.2 Encapsulate mock behavior

# Encapsulate mock behavior
def setup_mock_response(mock_get, status_code=200, json_data=None):
    """Configure the mock response.
    Args:
        mock_get (MagicMock): The mocked
requests.get
object.
        status_code (int): HTTP status code (default 200).
        json_data (dict): JSON payload to return (default None).
    """
    mock_response = MagicMock()
    mock_response.status_code = status_code
    if json_data is not None:
        mock_response.json.return_value = json_data
    if status_code >= 400:
        mock_response.raise_for_status.side_effect = requests.exceptions.HTTPError("Request failed")
    mock_get.return_value = mock_response

3.3 Write test cases

# Test case: successful request
def test_fetch_data_success(mock_requests_get):
    """Test
fetch_data
when the request succeeds."""
    setup_mock_response(mock_requests_get, status_code=200, json_data={"key": "value"})
    result = fetch_data("http://example.com/api/data")
    assert result == {"key": "value"}
    mock_requests_get.assert_called_once_with("http://example.com/api/data")

# Test case: failed request
def test_fetch_data_failure(mock_requests_get):
    """Test
fetch_data
when the request fails."""
    setup_mock_response(mock_requests_get, status_code=404)
    with pytest.raises(requests.exceptions.HTTPError):
        fetch_data("http://example.com/api/data")
    mock_requests_get.assert_called_once_with("http://example.com/api/data")

4. Run the tests

Execute the tests from the terminal:

pytest test_my_module.py

Explanation

Installation: pip install pytest installs pytest; unittest.mock is built‑in.

Module creation: fetch_data retrieves JSON from a URL and raises HTTPError on failure.

Test file: The fixture mock_requests_get patches requests.get .

Helper function: setup_mock_response configures status code and JSON payload for the mock.

Test cases: Verify correct behavior for both successful and error responses, and assert that requests.get is called with the expected URL.