DIG-1355: Sonchau/UUID with fk

[SonQBChau]

What's new:

• All tables will have FKs in the same manner as currently, but these FKs will be uuids
  submitter__ids will be stored in each object as a text field.
• UUIDs will remain internal to katsu and not be exposed via any API.
• UUIDs will be generated by katsu when data is ingested
• The combination of (program_id, submitter_*_id) will have a UNIQUE constraint in the database (for donor, specimen, sample, treatment and follow-up ids)

Potentially Breaking changes:

• DonorWithClinicalData no longer has filters. Can only retrieve 1 specific object by providing the program_id and submitter_*_id.
• The structure of the DonorWithClinicalData request has changed and is now "/v2/authorized/donor_with_clinical_data/program/{program_id}/donor/{donor_id}"
• Ingest APIs no longer accept a list of object.
• Discovery query no longer has filters. This will be addressed in the next patch
• New data ingest script for development using
  python chord_metadata_service/mohpackets/data/data_loader.py

Package changes:

• NEW django-ninja for both API and schema + filter
• NEW orjson faster JSON package for renderer since we use a lot of json
• REMOVE drf, replaced by ninja, used serializer which is slower and hard to customize
• REMOVE django-filter, replaced by ninja, only work with drf
• REMOVE drf-spectacular, replaced by ninja, only work with drf. This used to generate schema based on the serializer to yml file. Since the new ninja only generate json, we need to convert it to yml

How to test:

• Due to significant changes, it is recommended to drop the database and re-create
• pip install -r requirements/local.txt
• tox

Related tickets:

DIG-1389
DIG-1312

[daisieh]

As far as I can tell, this works as expected. Looks good!

[OrdiNeu]

From the data-portal side everything looks fine as far as I can tell, will need some modifications to data portal code to get it working with this but nothing too out-of-the-ordinary ✔️ .