This project demonstrates a way to test asynchronous Python code using a custom AsyncTestCase class. This approach is designed to work with Python 3.11 and later, where the asynctest.TestCase class is no longer supported.
Context: Stack Overflow question
In Python 3.10 and earlier, the asynctest.TestCase class was commonly used for testing asynchronous code. This class provided setUp and tearDown methods that supported the initialization of dependencies and creation of test data.
However, Python 3.11 and later versions do not support asynctest.TestCase. The unittest.IsolatedAsyncioTestCase class is an alternative, but it uses a different event loop for each test. This can lead to the usage of different sessions, which may not be desirable in some cases.
In this project, we define a custom AsyncTestCase class that extends unittest.TestCase. This class uses a single event loop for all tests, similar to asynctest.TestCase. It provides setUpClass and tearDownClass methods to initialize and clean up resources at the class level, and a run_async_test method to run asynchronous tests.
To run a test, define a test case class that extends AsyncTestCase and use the run_async_test method to run an asynchronous test. For example:
from tests.async_test_case import AsyncTestCase
from async_modules import async_add
class TestAsyncModule(AsyncTestCase):
def test_async_add(self):
async def coro():
result = await async_add(5, 3)
self.assertEqual(result, 8)
self.run_async_test(coro())