Removed dimension%nn3d and dimension%nn2d variable and replaced them with the actual dimensions kimax+1,kimax2+1

1. I had to introduce a temporary version of hmat due to issues related to
reading values that were already modified.
2. The input hmat had to be complex conjugated.

The changes were made to match the calculations of the old reference version.

...it is also increased in hybrid/hybrid.F90.

- Only half of the overlap matrix was set.
- invtrafo was missing a complex conjugation.

The requirement for these changes became obvious by a totalview comparison with
the old reference version. The values that are in the arrays match the old version
with very high precision.

Bugfixes for forces in spin-polarized case. Perhaps :-) closer to v26 now? Have to check this... #197

...it is not yet present.

Added a script to simplifly the generation of tests and modified the testing environment. Documentation still missing. Contributes to #190

Added explicit dimensions in calls to xcpot-routines to make code more stable. Fixed bug that should fix #191.

Bugfix in XC potential in MT spheres in cases with many atoms different jri and jspin=2. Should fix #193 @mhoff, please check.

Should fix #189. @zimmer please check!