Reimplement stores tests

[thclark]

Feature request

Current state

Stores were thoroughly tested before refactoring django-twined from another app. However, now, those tests are commented out and need to be brought up to scratch with the rest of the unit tests.

Also, stores require an integration test for various syncing purposes and we should use terraform to correctly hook up buckets for integration testing of the stores.

Note

I did a manual test of the stores, step-by-step evaluating what files appear and when, and when their metadata gets synced. The management command try_sync.py I used to do that is here:

from uuid import uuid4
from django.core.management.base import BaseCommand

from octue.resources import Datafile
from projects.tasks.trial import MyOnDemandTask
from icecream import ic


class Command(BaseCommand):
    """A management command to invoke on-demand and subscriber tasks manually
    Intended for use while experimenting in a live environment to see if it works
    """

    def handle(self, *args, **options):
        # An already existing file
        bucket_name = "test-lake-mast-timeseries"
        path = "77bff006-ca05-4384-a8a6-ff86e712d6ac.txt"
        datafile = Datafile(f"gs://{bucket_name}/{path}", id="77bff006-ca05-4384-a8a6-ff86e712d6ac")
        ic(datafile)
        ic(datafile.id)
        ic(datafile.labels)
        ic(datafile.tags)

        # Update metadata
        datafile.labels = ["stuff", "two"]
        datafile.update_metadata()

        ic(datafile.labels)
        ic(datafile.tags)
        ic(datafile.path_in_bucket)
        
        # A new file that shouldn't exist yet (we need to specify id, labels and tags then get metadata, all without creating a file

        creating_path = "whatever.txt"
        creating_id = str(uuid4())
        ic(creating_id)
        creating_df = Datafile(f"gs://{bucket_name}/{creating_path}", id=creating_id)
        ic(creating_df)
        # Now check - the creating_path should not be on gcs
        # Checked - correct
        ic(creating_df.metadata())
        # Now check - the creating_path should not be on gcs
        # Checked - correct
        creating_df.tags = {"a": "b"}
        ic(creating_df.metadata())
        # Now check - the creating_path should not be on gcs
        # Checked - correct
        creating_df.labels = ["a", "b"]
        ic(creating_df.metadata())
        # Now check - the creating_path should not be on gcs
        # Checked - correct