[GRASS-user] gdal - r.in.onearth

[Hamish]

Hamish:
> > G63> g.region -b
> > pj_transform() failed
> > cause: failed to load NAD27-83 correction file
> > ERROR: Error in pj_do_proj (projection of input coordinate pair)
> > 
> > that doesn't look good.

Markus:
> (as you will know)
> It indicates that 
>  http://proj.maptools.org/
>  -> proj-datumgrid-1.3.zip
>     ftp://ftp.remotesensing.org/proj/proj-datumgrid-1.3.zip
> 
> wasn't expanded in nad/ before compiling proj4.


The error was because I had changed the region to match a XY raster,
well outside of the location's nad grid file coverage.


### Spearfish60 location
G63> g.region rast=elevation.dem
G63> g.region -b
north longitude:  44:30:05.90947N
south longitude:  44:22:23.520696N
west latitude:    103:52:14.440271W
east latitude:    103:37:45.943809W
center latitude:  103:45:00.19204W
center longitude: 44:26:14.715083N

G63> g.region rast=xy2 -p
projection: 1 (UTM)
zone:       13
datum:      nad27
ellipsoid:  clark66
north:      1025
south:      -1025
west:       -1025
east:       1025
nsres:      50
ewres:      50
rows:       41
cols:       41
cells:      1681

G63> g.region -b
pj_transform() failed
cause: failed to load NAD27-83 correction file
ERROR: Error in pj_do_proj (projection of input coordinate pair)


I have seen this before when overlaying a geo-grid on a country-scale
zoom.

At minimum, the error message is misleading and should be changed.
It would be nicer if we could somehow know the extents of the nad file
and produce an out-of-region error instead.

r.in.onearth is not in GRASS CVS, otherwise about now I'd add a
if [ $? -ne 0 ] ; then
  echo "There was a problem"
  exit 1
fi
test just after the g.region call in the script.


Hamish