Make FDBM working with different codes

[ondrejkrejci]

Common issue accompanying these sub-issues:

• Suggest cleanup of geometry handling in the CPU version #198
• Figure out electrostatics in FDBM #277
• Grid geometry in OutFz.xsf not entirely correct #337
• Create a DFT example for FDBM testing from different codes #366
• Reading CHGCAR and POTCAR directly #342
• Enable loading XSF files from FHI-AIMS #217
• Wrong note on the *.xyz input format on Wiki #218
• Enable full input from Cube files, including densities #190

Suggested action:

@mondracek - before you will try to go deeply into writing your own reading procedure. ASE.IO has its reading procedure for cube and xsf together with the data. As our Choire is not anymore agains ASE as such, maybe it is worth of trying. The only thing that would be missing in that context is POTCAR/CHGCAR or AEC?CAR reader. Which is not even in ASE. If we would think about using ASE, then it would make sense to actually write it to ASE (I can handle the communication if necessary).