Allow the dispatcher to call handlers synchronously

[djalexd]

The current implementation will either store the dispatched event in event_store or create an asyncio.Task that is never awaited.

https://github.com/melvinkcx/fastapi-events/blob/dev/fastapi_events/dispatcher.py#L81

There are 2 situations where I believe synchronous handling is useful:

1. unit-testing. My current approach involves a somewhat complicated setup with a fixture:

@pytest.fixture(autouse=True)
def register_local_handlers():
    _id = 0
    token_middleware_id = middleware_identifier.set(_id)
    try:
        handler_store.update({_id: [local_handler]})
        yield
    finally:
        del handler_store[_id]
        middleware_identifier.reset(token_middleware_id)

this allows the registered local handlers to be invoked whenever I do a dispatch from tests

@pytest.mark.asyncio
async def test_save_success():
    with patch.object(DB, "save", new_callable=AsyncMock) as mock_save:
      dispatch("test-event", "test-payload")
     # here I cannot directly assert that handler performed something.
    
     # this is needed:
     await asyncio.gather(*asyncio.all_tasks())

2. having a handler that in turn forwards the event to some topic (i.e. SQS, SNS, Kinesis, Kafka), because this operation is very fast anyway. For example a retry policy would be hard to implement because there is no way I can await just for this dispatch task.

Any advice is welcome! I was also considering simply duplicating the dispatcher.py code as an alternative

[melvinkcx]

Thank you for bringing this up.

There are several solutions we can consider:

1. Introduce a set of test utils: Include a set of test utils that allows us to dispatch and handle events within test cases without writing custom fixtures, or even a function for us to trigger event handling right away.
2. Introduce another dispatcher or a new argument to dispatch() to allow us to dispatch an event and have it handled synchronously.

Let me know your thoughts?