`0x` orbitals compromise semicore selection logics

[npaulish]

This is weird, rare-earth pseudos have 0P/0D/0F orbitals, I check the UPF files and it is indeed there.
I am not sure what are these orbitals, as the code logics of ordering orbitals by principle quantum number would fail for these. I think we need to run some calculations to understand these pseudos, e.g. structures containing La element.

Originally posted by @qiaojunfeng in #74 (comment)