JetReconstruction: EDM improvements needed

[veprbl]

The current implementation of the JetReconstruction algo was done before JANA had native PODIO collections, so both jets and its constituents are stored in the same collection, with a hack:

• jets have type set to 0:
  

  EICrecon/src/algorithms/reco/JetReconstruction.cc

  Line 66 in 158e3e0

  jet_edm.setType(0);

• constituents have type set to 1:
  

  EICrecon/src/algorithms/reco/JetReconstruction.cc

  Line 80 in 158e3e0

  cst_edm.setType(1);

• jets are enumerated and PDG field used to store the index:
  

  EICrecon/src/algorithms/reco/JetReconstruction.cc

  Line 67 in 158e3e0

  jet_edm.setPDG(i);

• constituents refer to their jets by index, which is also stored in the PDG field:
  

  EICrecon/src/algorithms/reco/JetReconstruction.cc

  Line 81 in 158e3e0

  cst_edm.setPDG(i);

We first need to acknowledge that jets are not representable as edm4eic::ReconstructedParticle (e.g. they have "area"). Though even if they are for now, they don't belong in the same collection with the constituents. We also don't need to store constituents, instead we should just refer to the original ChargedReconstructedParticles/NeutralReconstructedParticles/MCParticles. That would not only allow to not duplicate the information unnecessarily, but also bring us closer to resolving a reported user issue: one currently can not relate jet constituents to the original track/cluster information.

cc @ruse-traveler @bspage912 @ShujieL

[veprbl]

See also #725