Factorize and improve modis reader's interpolation

[mraspaud]

This avoids recomputation of parallel datasets that are anyway needed for
the interpolation.

• Closes #xxxx
• Tests added and test suite added to parent suite
• Tests passed
• Passes flake8 satpy
• Fully documented

[djhoese]

Should these interpolation functions be delayed so that we aren't actually wasting memory by caching them?

[mraspaud]

From what I can see, the interpolation is dask-based, so I don't think a lot of memory is used

[djhoese]

Ah looks like you are right, nevermind then. LGTM.

[djhoese]

@mraspaud Is there anything else you are hoping to add to this PR or is it ready for review (it's still a draft currently)?

[mraspaud]

It's ready for review now

[djhoese]

Any idea how we get the travis tests to start up?

[mraspaud]

closing and reopening :(

[codecov]

Codecov Report

Merging #802 into master will increase coverage by 0.79%.
The diff coverage is 82.75%.

@@            Coverage Diff             @@
##           master     #802      +/-   ##
==========================================
+ Coverage   82.26%   83.06%   +0.79%     
==========================================
  Files         159      163       +4     
  Lines       22998    23592     +594     
==========================================
+ Hits        18920    19596     +676     
+ Misses       4078     3996      -82

Impacted Files	Coverage Δ
satpy/tests/reader_tests/test_modis_l2.py	97.43% <100%> (+0.02%)	⬆️
satpy/composites/__init__.py	70.55% <50%> (+1.31%)	⬆️
satpy/readers/hdfeos_base.py	88.48% <88.88%> (+9.09%)	⬆️
satpy/readers/aapp_l1b.py	13.65% <0%> (-0.21%)	⬇️
satpy/writers/utils.py	100% <0%> (ø)
satpy/tests/reader_tests/test_vaisala_gld360.py	96% <0%> (ø)
satpy/tests/writer_tests/test_utils.py	93.33% <0%> (ø)
satpy/readers/vaisala_gld360.py	90% <0%> (ø)
satpy/tests/test_scene.py	99.45% <0%> (ø)	⬆️
satpy/tests/reader_tests/test_virr_l1b.py	97.19% <0%> (+0.02%)	⬆️
... and 14 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6f1f451...778f078. Read the comment docs.

[coveralls]

Coverage increased (+0.8%) to 83.062% when pulling 778f078 on mraspaud:fix-modis-interpolation into 6f1f451 on pytroll:master.

[coveralls]

Coverage increased (+0.8%) to 83.062% when pulling 778f078 on mraspaud:fix-modis-interpolation into 6f1f451 on pytroll:master.