Refactor the usage of undefined function

[valentin-seitz]

This should fix #57.

I did test it in a Vlasiator installation i am doing and there applying these changes makes it compile with LLVM 19.

I didn't do a proper check if it semantically correct. I just went ahead and assumed that Types<3>::indices_t has the indices
stores as [x,y,z].

If you need some changes just let me know :)

[markusbattarbee]

We use Vlasiator-version of DCCRG which has diverged quite a bit, and probably won't be refactored together in the foreseeable future. So fixes should at least be made as PR's to that branch (thought I guess this can also be left for Ilja to consider!)

[valentin-seitz]

Hi @markusbattarbee,
As this "bug" is also present in the main branch I suspected it to be of interest for both versions.
If you want, i can just submit another PR targeting the vlasiator-version?

[markusbattarbee]

Well, I can't actually find the dimension-specific x/y/z_indices_overlap() functions in the Vlasiator-version branch at all, so I do somewhat wonder how you encountered this :)

[valentin-seitz]

Ahhh now i see!
The FIC branch i used for the analysis had an older submodule (https://github.com/fmihpc/dccrg/tree/cad44660c86a904ecf35061c92cc744e9ad4f37e) of DCCRG. I just checked out the vlasiator-version at HEAD and there its actually fixed :)
So i guess this is only interesting for master.