[WIP] Fix ADSIC potential
@wandongyu reported a weird behavior of ADSIC; the ADSIC gives different results between spin-polarized and unpolarized calculations even if the system is the closed shell. Then, I checked the code, and I found that the present implementation in Octopus is not consistent with the implementation of the original paper, J. Phys. B: At. Mol. Opt. Phys. 35 1115 (2002).
In the merge request, I fixed the code so that the octopus implementation is consistent with the paper.