let fre cmor run process tripolar ocean data

[ilaflott]

this PR will ideally enable fre cmor run to process ocean data on a native tripolar grid. also will do some general clean up here and there. relevant background: #268 PCMDI/cmor#766 and PCMDI/cmor#767

Describe your changes

• remove GitPython pip-only dependency that can only be fulfilled by terrible habits
• add PCMDI/cmip6-cmor-tables as submodule dependency
• set minimum python version to 3.9 minimum
• remove redundant click context forwarding in fre cmor
• remove redundant click context forwarding in most places, exception: fre catalog
• re-set os.environ['HOME'] back to inital value in test_fre_make_cli.py
• puts meta.yaml source: url: <this repo's URL> back to source: path: . in light of pipeline-error-induced insanity.
• pylint edits to address scary-ish no-value-for-parameter errors revealed by adjusting click scaffolding
• NotImplemented errors for fre functions that do not yet have any real guts to them in fre-cli
• get rid of sys.exit where possible
• removes "scratchwork" style fre cmor scripts used for by-hand local testing
• add's gfdl/ppan-only tests in fre/cmor/tests/test_cmor_run_subtool_ppan_only.py that gracefully fail when not at pp/an with an appropriate warning
• tweak pipeline comments
• make pylint kinder to long lines (--max-line-length 100 --> 120)
• remove unneeded __init__.py files in testing directories ... added back in, affected coverage stats
• last fre cmor run case of original 9 (Omon / sos / gn) case succeeds now

Issue ticket number and link (if applicable)

will close #298

Checklist before requesting a review

• I ran my code
• I tried to make my code readable
• I tried to comment my code
• I wrote a new test, if applicable
• I wrote new instructions/documentation, if applicable
• I ran pytest and inspected it's output
• I ran pylint and attempted to implement some of it's feedback

[ilaflott]

so asking conda/mamba to use python3.10 or greater is surprisingly an issue. Dinking around with this locally, the first thing it complains about is cylc-uiserver. Given that this has yet-to-be-utilized in the actual repository, I comment it out.

The next issue is catalogbuilder... actually, really, a requirement of catalogbuilder's. intake-esm==2023.7.7, which seems to desire a python of 3.9 or less, but 3.9 is nearing EOL.

Given that this quick python version bump was supposed to be a quick-sidequest, i'll be restoring the dependencies and barking up that tree some other time.

[ilaflott]

every now and then, a random conda build fails... here, it claims import git is causing the failure. I did remove that module, but i also removed that line

edit: same failure, same job. sanity demanding i ignore the conda_build workflow for a minute apparently

edit2: same failure, different job. xmas nightmare! moving on...

[ceblanton]

It looks good but admittedly I did not try it and review hard. In the future we will do more rigorous reviewing, promise!