ComponentGeoCoding from master

[SabineEmbacher]

No description provided.

[martin-boettcher]

There are many cosmetic changes and additions of a header that make it difficult to identify the changes relevant for the issue. I am therefore not sure I found all of them.
It seems to me the intention of the (accumulating) parameter pixelPos in functions that also return the pixelPos is that if we pass a parameter value here we will get it back. Else, a new one will be created. Since we pass one, the assignment afterwards is a duplication and should be avoided. Example: PixelGeoInverseIndex.getPixelPos() . There is at least one other location.

[marpet]

Yes, the issue was in the DistanceWeightingInterpolator.
21e48e3

Sometimes it changed the incomming PixelPos instance and sometimes it returned a new instance. In the commit above I made it consistent. In the Interpolator a reuse of the PixelPos instance is not intended. While it is for getPixelPos in InverseCoding. That's why the values need to be copied to the existing PixelPos.

[marpet]

@SabineEmbacher @martin-boettcher

Is this still releavant? Or can we clode the PR and delete the 8.x.calvalus_cgc branch?

[SabineEmbacher]

@marpet
Since this is already 2 years old, I think it is no longer relevant.

But this must bei decided by @martin-boettcher

[CLAassistant]

All committers have signed the CLA.