Feature/general pdos

[cpignedoli]

Contains also #189
This allows PDOS between two structures where the "molecule" part is not fully contained in the "full system part".
We should decide whether to integrate it in PDOS or to keep it separate.

[cpignedoli]

We need to allow a fragment to be an imported structure (e.g. from AiiDA database or ..).
We could even decide on flexibility with respect to the cell size of the fragment and offer the user the possibility to define "anchoring atoms"
We could add a viewer that is active for fragments and either visualizes the structure identified by the indexes of the fragment or visualizes a structure imported (from importers).

The fragment visualization should show the guessed constraints / spin guesses / .. and allow to modify them.
In case the fragment is an imported structure guesses should be empty or could be prefilled with the guess closest to the an that could be obtained from the "whole system"