JP-3546: fix deprecations and improve warning filtering

[emolter]

Resolves JP-3546

Closes #8296

This PR addresses replacing all deprecated code inside jwst, and replacing all instances of warnings.simplefilter("ignore") with more specific warnings like warnings.filterwarnings("ignore", category=RuntimeWarning, message="Divide by zero"). A few other easy-to-fix warnings were also fixed or suppressed.

There were two instances of warnings.simplefilter("ignore") which, when removed, did not create any additional warnings, making it difficult to track down the warnings that the filters were intended to handle. These were rules_level2_base.py line 565, and photom.py line 476. These filters were removed; however, it's possible that they should still be there, but the test coverage was inadequate to trip the warnings.

The Jenkins run was set up such that I could convince myself that only known and understood warnings would be skipped, and the rest would raise errors. I enforced this by modifying the ini_options in pyproject.toml. These changes to pyproject.toml should be reverted before approval! See the JIRA ticket for explanations of the warnings that are being filtered. Note that the oldest deps check is failing during parsing of these options, which will be removed anyway.

Jenkins run here

All regtest failures were due to stdatamodels version issues - failures are identical to run 1248. I'm sure there will be changes requested on this PR, and I'll plan to make a clean regtest run at that time.

Checklist for maintainers

• added entry in CHANGES.rst within the relevant release section
• updated or added relevant tests
• updated relevant documentation
• added relevant milestone
• added relevant label(s)
• ran regression tests, post a link to the Jenkins job below.
  How to run regression tests on a PR
• Make sure the JIRA ticket is resolved properly

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 77.32%. Comparing base (2fb073e) to head (df2b23f).
Report is 6 commits behind head on master.

❗ Current head df2b23f differs from pull request most recent head 61e90c1. Consider uploading reports for the commit 61e90c1 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #8320      +/-   ##
==========================================
+ Coverage   75.31%   77.32%   +2.01%     
==========================================
  Files         474      451      -23     
  Lines       38965    35089    -3876     
==========================================
- Hits        29345    27132    -2213     
+ Misses       9620     7957    -1663     

Flag	Coverage Δ		*Carryforward flag
nightly	77.32% <ø> (-0.02%)	⬇️	Carriedforward from 4cc0ac1

*This pull request uses carry forward flags. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[nden]

The regression test run linked above shows 12 warnings. A regular one shows ~500000. Is that right?

[hbushouse]

After reviewing again it looks good to me. Just a couple places where I think commented out lines could be removed.

[emolter]

The regression test run linked above shows 12 warnings. A regular one shows ~500000. Is that right?

@nden this is true, but misleading. Most warnings will return once the changes I made to the filterwarnings list in pyproject.toml are reverted. You can see the list of warning types I ignored for this PR there.

I tried to limit the scope of this PR to fixing DeprecationWarnings and to ensuring that all pre-existing calls to warnings.filterwarnings() were specific to the warning they intended to catch (so that e.g. no deprecation warnings were hidden). I did not attempt to remove all warnings in the code; this would be a much bigger project, although a few other small bits of warning-related cleanup did end up in here, or triggered by this (e.g. JP-3555)

[nden]

LGTM