Minimum change to reflect new bucket locations

[metazool]

See #12 for context (changes to the layout of the object store).

• Updates to the bucket locations (separate collections for FlowCam and flow cytometer images)
• intake metadata and catalog (just a file listing) have moved into the image buckets alongside the cyto_app metadata
• Adds a basic test that checks files can be read from the new buckets, only runs if connection details are set in .env

This is a small and quite blunt change - i think any more work on making it generic should defer until

• Revisit how the "decollage" process works in cyto_app so that the spatio-temporal metadata that is available with the sources is preserved along with the individual images
• Have a much closer look at the flow cytometer data - if in theory we can get spectrograms out of it as well, then just pushing the images through a classifier model is losing potential...

To test

pip install -e .
py.test 

The tests for new bucket locations will skip unless you have ENDPOINT and the JASMIN access tokens set in .env (they should skip in the pipeline but that's ok imo, just doing a slow file listing)

@Kzra

[metazool]

Considered adding documentation for the approach to intake here (expected to really complete #12) but better to keep PRs small and actually finish them :D

[github-actions[bot]]

☂️ Python Coverage

current status: ✅

Overall Coverage

Lines	Covered	Coverage	Threshold	Status
123	105	85%	0%	🟢

New Files

File	Coverage	Status
cyto_ml/tests/test_object_store.py	60%	🟢
TOTAL	60%	🟢

Modified Files

File	Coverage	Status
cyto_ml/data/s3.py	64%	🟢
cyto_ml/tests/conftest.py	100%	🟢
TOTAL	82%	🟢

updated for commit: 73d5c1e by action🐍

[metazool]

Updated to fix a test failure where .env is present but with arbitrary string instead of ENDPOINT location

Should bring this back in sync with the state of cyto-ML project @Kzra - are you happy to add an approving review? Please also see an issue there about preserving more of the metadata during the decollaging process :D

[metazool]

Much appreciated, @albags !