Generalize NCPositivityCharge

[alecandido]

@felixhekhorn introduce this apparently over-specific setting. But at the moment is really fitting the purpose of what we needed.

A possible generalization is to extend it, in a more straightforward way, to:

• a list
• of (absolute values of) PIDs of particles
  with non-zero coupling to the electro-weak boson (absolute values needed: it would not make sense to couple particles and not anti-particles).

This require something a bit more involved also in the logic, since EM and NC can filter over a single flavor, but CC have to take into account 2 flavors at a time (and they have both to be included in the list to pass the selection).