Excise Truth Info from Electron Finder, Prepare for Calorimeter PID

[ruse-traveler]

Currently, the 1st stage of the electron finder (ElectronReconstruction) implicitly makes use of truth information by way of its input, the output of the MatchClusters algorithm. An important step will be removing this dependence on truth information.

Furthermore, a point discussed during the July 2024 UGM eID/Reconstruction workfest was that the existing ElectronReconstruction algorithm could be evolved into a more general calorimeter-based lepton/hadron discriminator.

The former point can addressed while laying the groundwork for the latter by wiring in a judicious choice of inputs/outputs (c.f. edm4eic#97).

Describe the solution you'd like
A natural choice of inputs would be the TrackClusterAssociations of edm4eic#52 produced by a centralized track-cluster matching algorithm (c.f. EICrecon#1648). Then, by having the ElectronReconstruction output ChargedRecoParticleCandidates would allow it to be wired into more general PF routines, functioning as a lepton/hadron discriminator.

Describe alternatives you've considered
Alternatively, the truth information can be excised by replacing MatchClusters with a more suitable PF algorithm. However, this is less desirable as the existing outputs might make it more challenging to evolve into the lepton/hadron discriminator.

Additional context
This falls into the Reconstruction WG's eID/PFA development program.

[veprbl]

Truth info needs to be now banished from the CalorimeterParticleIDPreML and from MatchClusters. Both CalorimeterParticleIDPreML and CalorimeterParticleIDPostML probably need to start consuming ReconstructedParticles and pick info from related (rename those algorithms then).