wsgiref-fake

In-process WSGI test doubles: fake sockets and a minimal server harness so you can exercise HTTP (and optionally WebSocket) applications without binding a real port or talking over the network.

Useful for fast, deterministic tests against code built on wsgiref and similar stacks.

Requirements

• Python 3.10 or newer
• Runtime dependency: six (Python 2/3 compatibility helpers used in this library)

Installation

pip install wsgiref-fake

With uv:

uv add wsgiref-fake

Optional extras

pip install "wsgiref-fake[gevent]"
pip install "wsgiref-fake[ws4py]"

Quick start (HTTP)

Wire your WSGI app to an in-process server, then drive it with HTTPClient:

from wsgiref_fake.http.client import HTTPClient
from wsgiref_fake.server import make_server


def app(environ, start_response):
    start_response("200 OK", [("Content-Type", "text/plain")])
    return [b"Hello World"]


server = make_server(app=app)
client = HTTPClient(server=server)

response = client.request(method="GET", path="/")
assert response.status == 200
body = response.read()

HTTPClient.request accepts method, path, optional json (sets Content-Type and body), and optional headers.

Why this exists

• No listening socket — tests stay hermetic and parallel-friendly.
• Same WSGI surface as production-oriented wsgiref servers, without TCP overhead.
• WebSocket-related helpers are available when you install the optional dependency groups above.

Development

Clone the repository, then install the project with dev dependencies:

uv sync

Run the test suite:

uv run pytest

Lint and format the whole tree:

uv run ruff check .
uv run ruff format .

Contributing

Issues and pull requests are welcome. Please run tests before opening a PR.

• Issue tracker
• Source code

License

This project is licensed under the MIT License.