Draft: Redoing charm builders in BandQ doublecharm lines

Implementing suggestion from !4070 (comment 9006064) (New MR from !4475 (closed) to avoid picking up diverging commits)

We should carefully consider if one really needs to make this change. And also evaluate it's effect on both rates and at least D0 D0 efficiency. But at least the draft is here.

• switch from using uncontrolled makers from CharmWG to in-house make_charm_to_hadrons from c_to_hadrons.py, while putting ownpvipchi2 < 16 as an additional requirement on these charm hadrons.

• also redone list of decay_descriptors in semi-automatic way thus

  • ensuring number of decay modes "missed" previously: like J/psi(1S) J/psi(1S) and other PsiPsi, and all with Xi_c+
  • and changing descriptors from anti-charm charm to charm anti-charm

• merged together samesign and opposite sign. Why to keep them separate?