wsgitest

You can use wsgitest to start a HTTP server for a WSGI application and control the process:

@Request.application
def application(request):
    return Response('Hello World!')

class AppServerTestCase(TestCase):
    def test_hello_app(self):
        server = WSGITestServer.create(application)
        try:
            response = requests.get(server.application_url)
            self.assertEqual(response.status_code, 200)
            self.assertEqual(response.content, b"Hello World!")
        finally:
            server.terminate()

    def test_hello_app_reference(self):
        server = WSGITestServer.create("tests.test_server.application")
        try:
            response = requests.get(server.application_url)
            self.assertEqual(response.status_code, 200)
            self.assertEqual(response.content, b"Hello World!")
        finally:
            server.terminate()

The method WSGITestServer.create() initialize a multiprocessing.Process and wait for the server startup.

You can use server as a context manager to avoid that tests get stucked when you forget to "terminate" server:

@Request.application
def application(request):
    return Response('Hello World!')

class AppServerTestCase(TestCase):
    def test_hello_app(self):
        with WSGITestServer(application) as server:
            response = requests.get(server.application_url)
            self.assertEqual(response.status_code, 200)
            self.assertEqual(response.content, b"Hello World!")

You can also start server with an specific host/ip or TCP port:

@Request.application
def application(request):
    return Response('Hello World!')

class AppServerTestCase(TestCase):
    def test_hello_app(self):
        with WSGITestServer(application, "0.0.0.0", 5000) as server:
            response = requests.get(server.application_url)
            self.assertEqual(response.status_code, 200)
            self.assertEqual(response.content, b"Hello World!")