Upstream 2024 04 22

[EddieLF]

Pulls in new upstream changes from after 2024-01-15 (#198), up until 2024-04-22.

Merge conflicts came up in two main areas:

• Anvil / Airtable code from upstream which has mostly been commented out for our deployment. I've pulled in the latest changes made upstream and continued with commenting out the same blocks of code.
• Cas' changes regarding AIP reports and tagging. I resolved merge conflicts by preferring Cas' changes from staging, all of which were newer than the upstream code.
• Version updates in requirements. I favoured newer versions when resolving these conflicts, although I did check if there were any versions of any modules we had pinned for some reason, I didn't see anything specifically pinned being overwritten.

requirements.txt were rebuilt after pulling in changes with pip-compile requirements.in.

---

Some noteworthy changes from the changelog

• Add subscribable project notifications (REQUIRES DB MIGRATION)
• Add "Probably Solved" analysis status (REQUIRES DB MIGRATION)
• Support FRASER2 data (REQUIRES DB MIGRATION)
• Add solve_status to Individual model (REQUIRES DB MIGRATION)
• Improved Variant Lookup

[EddieLF]

@illusional I think I've got it down to one failing python unit test, in the dataset_api_tests.py where it's mocking an email. It mocks the email with an f-string containing the {SEQR_URL} global defined at the top of the file. For some reason the URL string is missing from the email content when the assert gets called. The process_message is also different.

I'm confused why this test is failing, where there are a bunch of other tests that make use of _assert_expected_notificationwhich aren't failing. Any ideas?

[illusional]

Nice @EddieLF! I imagine it comes from here: https://github.com/populationgenomics/seqr/blob/8731e2f16113bf7da189a60bd9e15b568235f3bb/seqr/utils/search/add_data_utils.py#L58 (which comes from) https://github.com/populationgenomics/seqr/blob/8731e2f16113bf7da189a60bd9e15b568235f3bb/settings.py#L312

Up to you if you want to modify the https://github.com/populationgenomics/seqr/blob/8731e2f16113bf7da189a60bd9e15b568235f3bb/seqr/views/apis/dataset_api_tests.py#L14 in the test to just be a /, or add an environment variable to the tests (I'd probably lean towards the former, because the testing environment is simpler, and tests the code, not the configuration).

Or mock the variable in the test to be what's listed above, is also totally fine too.

[illusional]

@EddieLF, just following up because I could test it locally, that function is only called within that one test (and it's failing on the first invocation), and it was added in this merge.

[EddieLF]

@illusional thanks for taking a look and suggesting a fix. I've added a @mock.patch for the BASE_URL in the add_data_utils to mock the existence of the environment variable. This has resolved the test failure, however overall the Python unit tests are still marked as failing with exit code 2. I took a look through the logs and nothing jumped out, all pytest modules report ok and cmd+f fail turns up nothing. There are some warnings about "DateTimeField %s received a naive datetime (%s)", but these don't indicate failure. Any idea what I might be missing? Cheers

[illusional]

It's failing because the codecoverage fell under 90% (in the unittest). If I were you, I'd just drop the minimum coverage (in the unit-test.yaml) to like 85%.