eclipse_conical issue

[kevin-tracy]

There is an issue with eclipse_conical. Here is a simple example where eclipse_cylindrical gives the expected result but eclipse_conical does not:

julia> eclipse_cylindrical(R_EARTH*1.2*[1;0;0],AU*[1;0;0]) # should be fully illuminated (1.0)
1.0

julia> eclipse_conical(R_EARTH*1.2*[1;0;0],AU*[1;0;0])     # should be fully illuminated (1.0)
0.0

This isn't caught by the tests because it looks like the expected illuminations are flipped here: https://github.com/sisl/SatelliteDynamics.jl/blob/master/test/test_orbitdynamics.jl#L107. I'll submit a PR this weekend with the fix for this, but I figured I would add an issue anyways.