Hi all,
looking at the LiteRenderer/GridCoverageRenderer code I noticed that the latter doesn't take care of the possible difference between the coverage CRS and the renderer/display CRS.
Building a GridCoverage2D means basically build a WritableRaster and a GridGeometry consistent with the CRS associated to the GridCoverage itself. So for example if we have a WGS84 CRS with (NORTH,EAST) axis associated to the GridCoverage2D, that means we have a (row,column) coordinate system with the x and y axis increasing, we have to create an AffineTransform (aka gridToCoordinateSystem transform) which swaps the two axis and reverses the y one. Supposing that the renderer has a (EAST,NORTH) CRS, as ususal, in the case we don't care of the difference between the two CRSs, that is the actual situation of the LiteRenderer, at the end we will have a rendered image with the two axis wrong swapped.
Another problem, I think, is that it's not sufficient re-swap the two axis before rendering because we have even to recalculate the scale an offset against the screen dimensions (aka target grid-range).
Is it right what I'm saying here, or I'm missing something?
How can I take care of those two problems?
Is there someone alredy working on that?
Thank you very much.
Cheers,
Alessio.
