Introduction ¶

The article presents an alternative to the requests_mock library for testing HTTP requests in Python. The author demonstrates why “responses” library offers simpler usage with more advanced features. As noted, mocking external service calls prevents test failures when services are unavailable and eliminates unnecessary costs during testing.

Installation ¶

Simple installation via: pip install responses

Basic Usage ¶

The author illustrates replacing requests_mock with responses using a weather subscription test case. The decorator @responses.activate enables mocking functionality, followed by configuring mock responses:

responses.add(
    method=responses.POST,
    url='https://real-weather-service.com/weather/subscribe/',
    json=return_value,
    status=201
)

Request Matching ¶

The library supports eight matching strategies:

• JSON body contents
• URL-encoded data
• Query parameters
• Query strings
• Request kwargs (stream, verify)
• Multipart form-data content
• Headers
• Fragment identifiers

A practical example demonstrates URL-encoded parameter matching, with a critical note: parameters must be converted to strings for proper matching, as URL encoding converts all values to strings.

Dynamic Responses ¶

Using callbacks enables conditional response logic. Callbacks examine request details and return tuples of (status, headers, body):

def request_callback(request):
    payload = dict(parse_qsl(request.body))
    # Process and return response
    return (status, headers, json.dumps(resp_body))

responses.add_callback(
    method=responses.POST,
    url='https://...',
    callback=request_callback
)

Conclusion ¶

The author concludes that “responses” provides more sophisticated features with greater ease of use compared to alternatives, making it the preferred choice for request mocking in Python testing.