Add a reader for MODIS Level 3 files in CMG format.

[simonrp84]

This PR adds a reader for MODIS Level 3 (daily / monthly / etc) products on the climate monitoring grid.

I'd like feedback on the change I made to hdfeos_base.py. The CMG data has metadata containing URLs with = signs, which confuses the _split_line method. My change means that only the first = is used to split strings, with the subsequent ones assumed as part of the metadata.
Personally, I don't think the try...except is needed as there should only ever be one important = in each string. But I wrote it this way in case there's some problem I've not thought of. Opinions welcome.

• Tests added
• Fully documented

[codecov]

Codecov Report

Attention: 3 lines in your changes are missing coverage. Please review.

Comparison is base (6fb1d1e) 95.20% compared to head (0a18d2d) 95.22%.
Report is 12 commits behind head on main.

Files	Patch %	Lines
satpy/readers/modis_l3.py	96.36%	2 Missing ⚠️
.../tests/reader_tests/modis_tests/_modis_fixtures.py	97.91%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2631      +/-   ##
==========================================
+ Coverage   95.20%   95.22%   +0.01%     
==========================================
  Files         354      356       +2     
  Lines       51370    51539     +169     
==========================================
+ Hits        48908    49076     +168     
- Misses       2462     2463       +1     

Flag	Coverage Δ
behaviourtests	4.22% <0.00%> (-0.02%)	⬇️
unittests	95.84% <98.17%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[coveralls]

Pull Request Test Coverage Report for Build 6892910354

Warning: This coverage report may be inaccurate.

We've detected an issue with your CI configuration that might affect the accuracy of this pull request's coverage report.
To ensure accuracy in future PRs, please see these guidelines.
A quick fix for this PR: rebase it; your next report should be accurate.

• 161 of 164 (98.17%) changed or added relevant lines in 6 files are covered.
• 8 unchanged lines in 4 files lost coverage.
• Overall coverage increased (+0.01%) to 95.795%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
satpy/tests/reader_tests/modis_tests/_modis_fixtures.py	47	48	97.92%
satpy/readers/modis_l3.py	53	55	96.36%

Files with Coverage Reduction	New Missed Lines	%
satpy/tests/modifier_tests/test_angles.py	1	99.5%
satpy/tests/reader_tests/modis_tests/_modis_fixtures.py	1	98.23%
satpy/modifiers/angles.py	2	99.22%
satpy/readers/nwcsaf_nc.py	4	82.21%

Totals
Change from base Build 6847297703:	0.01%
Covered Lines:	49202
Relevant Lines:	51362

---

💛 - Coveralls

[simonrp84]

Codefactor is moaning about the complexity of create_hdfeos_test_file and I'm not really sure what I'm supposed to do about it...all the if statements there are needed and it seems kinda wasteful to invent a whole new function just to avoid if..then.else. Suggestions appreciated!

[djhoese]

Ok so this is just assuming a generic lon/lat coordinate system?

https://gdal.org/user/coordinate_epoch.html

Looks like a related discussion:

opengeospatial/geoparquet#52

[djhoese]

I think your suggestion on slack to use OGC:CRS84 is good as a minimum. I could have sworn there was a CRS that was at least rooted in an exact date/time, but otherwise was similar to EPSG 4326. At least CRS84 has the same axis order that we use/expect.

[djhoese]

CodeScene is happy now!

[djhoese]

Nice job putting this together. I put a couple questions, some concerns, and other comments inline.

Otherwise, I'd like to officially state that I'd like something better than the -999/-9999 resolution stuff in the tests to describe these files. It just seems so hacky. If it must be this way then fine, but it feels like we can come up with something better.

[djhoese]

This looks good to me. I wouldn't mind a discussion on the usage of EPSG:4326 and I'd still like @mraspaud to look at the change to hdfeos_base.py, but otherwise this looks pretty good. Nice job sticking with it @simonrp84!

[simonrp84]

Looking through the MODIS docs it seems like EPSG:4326 is what's used by the data provider, so I suggest that we stick with that in this PR.

[mraspaud]

Looks good, but I would use the maxsplit everytime.

[simonrp84]

Looks good, but I would use the maxsplit everytime.

Sorted this, so always using maxsplit=1 now.

[djhoese]

I'll see if I can do a quick check on the viirs edr reader so complete the tests for these changes. Then we can merge it.

[simonrp84]

That was a monster of a PR, turned out to be more complex than I was anticipating! Thanks for the help and comments along the way, really nice to get this merged.