problem in grid mapping prj

[mdsumner]

@dblodgett-usgs I see a problem in this test test-gridmapping-prj on Windows, but not Linux.

It comes down to this including "semi_minor_axis" on Windows but "inverse_flattening" on Linux.

p <- "+proj=lcc +lat_1=30 +lat_2=60 +x_0=0 +y_0=0 +units=m +lat_0=40.0000076294 +lon_0=-97 +a=6370000 +b=6370000 +pm=0 +no_defs"
nc_prj_to_gridmapping(p)

LINUX:

# A tibble: 9 x 2
  name                          value    
  <chr>                         <list>   
1 grid_mapping_name             <chr [1]>
2 standard_parallel             <dbl [2]>
3 false_easting                 <dbl [1]>
4 false_northing                <dbl [1]>
5 latitude_of_projection_origin <dbl [1]>
6 longitude_of_central_meridian <dbl [1]>
7 semi_major_axis               <dbl [1]>
8 inverse_flattening            <dbl [1]>
9 longitude_of_prime_meridian   <dbl [1]>

WINDOWS:

# A tibble: 9 x 2
  name                          value    
  <chr>                         <list>   
1 grid_mapping_name             <chr [1]>
2 standard_parallel             <dbl [2]>
3 false_easting                 <dbl [1]>
4 false_northing                <dbl [1]>
5 latitude_of_projection_origin <dbl [1]>
6 longitude_of_central_meridian <dbl [1]>
7 semi_major_axis               <dbl [1]>
8 semi_minor_axis               <dbl [1]>
9 longitude_of_prime_meridian   <dbl [1]>

But I wasn't able to pin it down any further yet (I did make some changes in the the GGFP methods to make the return value explicit).

[dblodgett-usgs]

Huh... I'm not surehow that test was ever passing that way! PR on the way.