Fix failing tests

[haatveit]

While trying to switch from the manual test rigging to using pytest (ae4b1e3) in branch tests-cleanup, this uncovered some issues with the currently burrowed code in previously inaccessible functions that have probably not seen much execution.

We should rewrite or drop the tests if they're outdated, and fix the underlying issues in case any of them are symptoms of bugs/regression.

Fix failing tests

Beta Give feedback

1. test_XXX_load
2. test_ess_migration
3. test_ZZc_requesting_multiple_ranges_supported (regression?)
4. test_ZC_setting_ownership_based_on_user_works
5. test_X_use_generic_table

• Both test_XXX_load and test_ess_migration require postgres but do not currently seem to pass even when a database is present in CI, need more investigation. Running against postgres locally requires a bit of effort from me so I haven't gotten there yet.
  • Maybe test_ess_migration can go either way?
• test_ZZc_requesting_multiple_ranges_supported failure points to a bug or regression introduced after the test was initially added, as specifying multiple ranges currently leads to weird behavior. We will need to look at how the ranges are parsed.
• test_ZC_setting_ownership_based_on_user_works does not work with the default configuration of tsd-file-api, will need configuration file or new defaults (probably the test should confirm the require option(s?) is set in the file API).
• test_X_use_generic_table I've just told pytest to skip for the time being, I think it uses an old endpoint that's not supported anymore?