Fix for duplicate coordinates in bbox_lonlats for geostationary area

[ghiggi]

This PR removes the presence of duplicate consecutive coordinates in the geostationary are bbox/polygon.
This currently leads to the creation of SArcs that are actually points, and also leads to cross-product operations that lead to cartesian [0,0,0] and thus vector norm = 0 and thus division by 0 when normalizing.

Furthermore, it closes #474: when we retrieve 4 vertices we now get the 4 correct geos_area "extent" coordinates.

• Closes get_geostationary_bounding_box* contains duplicated vertices at the equator  #474
• Tests added
• Tests passed
• Passes git diff origin/main **/*py | flake8 --diff

[codecov]

Codecov Report

Merging #475 (d22496f) into main (1268d90) will decrease coverage by 0.04%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #475      +/-   ##
==========================================
- Coverage   94.28%   94.24%   -0.05%     
==========================================
  Files          69       73       +4     
  Lines       12394    12703     +309     
==========================================
+ Hits        11686    11972     +286     
- Misses        708      731      +23     

Flag	Coverage Δ
unittests	94.24% <100.00%> (-0.05%)	⬇️

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

Impacted Files	Coverage Δ
pyresample/geometry.py	87.17% <100.00%> (ø)
pyresample/test/test_geometry.py	99.51% <100.00%> (ø)
pyresample/spherical.py	92.67% <0.00%> (-5.55%)	⬇️
pyresample/test/test_spherical.py	100.00% <0.00%> (ø)
pyresample/future/spherical/point.py	100.00% <0.00%> (ø)
pyresample/future/spherical/__init__.py	100.00% <0.00%> (ø)
pyresample/test/test_sgeom/__init__.py	100.00% <0.00%> (ø)
pyresample/test/test_sgeom/test_point.py	100.00% <0.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[coveralls]

Coverage decreased (-0.03%) to 93.761% when pulling d22496f on ghiggi:bugfix-duplicate-geo-bbox-lonlats into 1268d90 on pytroll:main.

[djhoese]

I think this makes sense to the extent (pun intended) that I understand it. The solution seems simple enough. Thanks for fixing this. I'll merge it once tests pass and you can merge it into your boundary PR #473 and then we can merge that.

[djhoese]

Some of the uncovered lines that coveralls is complaining about are a little concerning, but I assume you'll be working with this in the upcoming PRs so I'm not too worried. Most of the missed lines are the matplotlib plotting code which is annoying to test anyway. The one concerning one is the vertices to xyz helper function. The branch with the .T is reported as uncovered. Let's merge this for now and have you work on covering that line in one of the other PRs.