Inconsistent conversions of LatLng and MGRS coordinates

[BigBallard]

I am having some troubles with conversations between LatLng and MGRS providing inconsistent and incorrect values for some geographic areas in different hemispheres and also within those hemispheres being inconsistent. To help verify I provided some the test points I used to check the validity of the conversions:

NW Hemisphere:

• Kansas USA (Consistent):
  • Lat Lng: 39.964463, -99.820180 -> MGRS: 14SMK2995324135
  • MGRS: 14SMK2995324135 -> Lat Lng: 39.964460, -99.820180

SW Hemisphere:

• Rio De Janeiro Brazil (Inconsistent)
  • Lat Lng: -22.938733, -43.503667 -> MGRS: 23KPQ5342862481
  • MGRS: 23KPQ5342862481 -> Lat Lng: -76.724974 -35.068682

NE Hemisphere:

• Leipzig Germany (Consistent)
  • Lat Lng: 51.349262, 12.365591 -> MGRS: 33UUS1655091960
  • MGRS: 33UUS1655091960 -> Lat Lng: 51.349262, 12.365591
• Kassel Germany (Inconsistent)
  • Lat Lng: 51.310206, 9.536331 -> MGRS: 32UNB3738384458
  • MGRS: 32UNB3738284458 -> Lat Lng: 65.679762, 9.813488

The code I am using the create and convert LatLng to MGRS:

double lat = 39.964463;
double lon = -99.820180;
LatLng ll = new LatLng(lat, lon);
MGRSRef mgrs = ll.toMGRSRef();

And then building off that, how I convert MGRS to LatLng by taking the string value of the MGRS, create a new MGRSRef , and convert that to LatLng:

MGRSRef mgrs = new MGRSRef("14SMK2995324135");
... = mgrs.toLatLng()

[xni06]

Hi @DallasP9124 - I can't help sorry but feel free to add a PR or take a look at the forks to see whether anyone else has made sense of it at all.

I've just taken the author's source and made it available to the community