Reorder Hubbard atoms with HubbardStructureData

[bastonero]

The hp.x code will crash if the hubbard atoms are not listed before all the other ones. Thus, at the beginning of the SelfConsistentHubbardWorkChain we should check for this. The HubbardStructureData is already provided with a method that internally will reorder the sites.

We need simply to implement a simple calcfunction that takes as input a HubbardStructureData, clones it, uses the self.reorder_hubbard_atoms(), and gives in outputs an ordered HubbardStructureData.

This functionality was already implemented before the HubbardStructureData implementation. The method will make these procedure smoother.

[sphuber]

This functionality was already implemented before the HubbardStructureData implementation. The method will make these procedure smoother.

Indeed, see this calcfunction: https://github.com/aiidateam/aiida-quantumespresso-hp/blob/master/src/aiida_quantumespresso_hp/calculations/functions/structure_reorder_kinds.py

[bastonero]

Thanks @sphuber ! Yes, I new about that calcfunction. Now that we are using the HubbardStructureData, also the Hubbard parameters themselves must be reordered accordingly within the class. This feature has been already implemented in the dedicated branch on aiida-quantumespresso.

IMPORTANT to discus: shall we "pre" reorder the sites within the HubbardStructureData "on storing"? I think no, since pw.x runs even without this order.

[bastonero]

Fixed in #32