Preserve DATE-BEG, DATE-END when creating level2 files

[d-lamb]

PR summary

When creating observatory-specific L2 files from L1 files, make sure the beginning and ending dates (DATE-BEG and DATE-END) are preserved in the L2 files. Also, when creating mosaic files, use the earliest input DATE-BEG and the latest input DATE-END in the mosaic.

This is one possible approach. We could adopt a different approach in which the earliest and latest DATE-OBS is what goes into the DATE-BEG and DATE-END of the mosaic.

Todos

Please list tasks to complete this PR.

• Verify that the new behavior is what is actually wanted
• Make sure that it also works for L3 PTM where 3 L1 polarized files go into the L2 XP files
• Verify new tests pass
• Verify that changes make it all the way to L3 CTM and L3 PTM (This can be done when running a full deployment test.)
• Address the file provenance issue in comment to L3_CTM files have all date fields set the same #948. (Upon further reflection, I think this can be skipped since the L3_CTM files don't directly inherit from the L2_CTM files, but rather from the L2_X files.)

Test plan

Added many tests to auto/level2

Breaking changes

None known

Related Issues

Closes #948

[codecov]

Codecov Report

❌ Patch coverage is 93.58974% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 58.24%. Comparing base (280159b) to head (ce41dec).
⚠️ Report is 44 commits behind head on main.

Files with missing lines	Patch %	Lines
punchbowl/data/meta.py	77.77%	4 Missing ⚠️
punchbowl/auto/flows/level3.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #963      +/-   ##
==========================================
+ Coverage   58.04%   58.24%   +0.19%     
==========================================
  Files         127      127              
  Lines       13833    13952     +119     
==========================================
+ Hits         8030     8126      +96     
- Misses       5803     5826      +23     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[d-lamb]

This is close, but not quite there. Currently it is doing a good job of settting the database fields from auto/flows, but needs to set the same info in the file header in level2/merge.py.

[d-lamb]

OK, @jmbhughes, @svank, @lowderchris this is ready for review.

[lowderchris]

I wonder if it would help testing to make one of the WFIs late, or something - to check how that will change things downstream in a predictable way.

[d-lamb]

@lowderchris The commit ce41dec makes the two banks of tests the same, replaces the chained equality (which was cumbersome for 4 CTM files) with the len(set()) idiom, and actually checks the values of the returned date_beg, date_obs, and date_end, which is what I think you were going for here.

[lowderchris]

This is looking great! Time at last... I just left a suggestion or two and a question.