replayer_storage_and_client = (<swh.storage.in_memory.InMemoryStorage object at 0x7f50f7258b00>, <swh.journal.client.JournalClient object at 0x7f50f9359940>) caplog = <_pytest.logging.LogCaptureFixture object at 0x7f50f93929e8> def test_storage_replayer(replayer_storage_and_client, caplog): """Optimal replayer scenario. This: - writes objects to a source storage - replayer consumes objects from the topic and replays them - a destination storage is filled from this In the end, both storages should have the same content. """ src, replayer = replayer_storage_and_client # Fill Kafka using a source storage nb_sent = 0 for object_type, objects in TEST_OBJECTS.items(): method = getattr(src, object_type + "_add") method(objects) if object_type == "origin_visit": nb_sent += len(objects) # origin-visit-add adds origin-visit-status as well nb_sent += len(objects) caplog.set_level(logging.ERROR, "swh.journal.replay") # Fill the destination storage from Kafka dst = get_storage(cls="memory") worker_fn = functools.partial(process_replay_objects, storage=dst) nb_inserted = replayer.process(worker_fn) assert nb_sent == nb_inserted > _check_replayed(src, dst) .tox/py3/lib/python3.7/site-packages/swh/storage/tests/test_replay.py:87: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ src = <swh.storage.in_memory.InMemoryStorage object at 0x7f50f7258b00> dst = <swh.storage.in_memory.InMemoryStorage object at 0x7f50f1772710> exclude = None def _check_replayed( src: InMemoryStorage, dst: InMemoryStorage, exclude: Optional[Container] = None ): """Simple utility function to compare the content of 2 in_memory storages """ expected_persons = set(src._persons.values()) got_persons = set(dst._persons.values()) assert got_persons == expected_persons for attr in ( "contents", "skipped_contents", "directories", "revisions", "releases", "snapshots", "origins", "origin_visits", ): if exclude and attr in exclude: continue expected_objects = sorted(getattr(src, f"_{attr}").items()) got_objects = sorted(getattr(dst, f"_{attr}").items()) > assert got_objects == expected_objects, f"Mismatch object list for {attr}" E AssertionError: Mismatch object list for origin_visits E assert [('https://ov...}, visit=3)])] == [('https://ov...den/fox', [])] E At index 0 diff: ('https://overtherainbow.org/fox/den', [OriginVisit(origin='https://overtherainbow.org/fox/den', date=datetime.datetime(2014, 11, 27, 17, 20, 39, tzinfo=tzlocal()), status='ongoing', type='hg', snapshot=None, metadata={'baz': 'qux'}, visit=1), OriginVisit(origin='https://overtherainbow.org/fox/den', date=datetime.datetime(2015, 11, 27, 17, 20, 39, tzinfo=tzlocal()), status='partial', type='hg', snapshot=b'\xec\xeeH9z\x92\xb0\xd04\xe9u*\x17E\x9f6\x91\xa7>\xf9', metadata={'something': 'wrong occurred'}, visit=2)]) != ('https://overtherainbow.org/fox/den',... E E ...Full output truncated (13 lines hidden), use '-vv' to show .tox/py3/lib/python3.7/site-packages/swh/storage/tests/test_replay.py:209: AssertionError