[Halld-offline] reorganization of tracking and pid code

[Simon Taylor]

Hello, all.

Based on the discussion we had at the most recent offline meeting, I 
have reorganized the tracking code in the following manner:  at the 
wire-based fitting stage, each track candidate is fitted multiple times 
based on a list of mass hypotheses.  All tracks that survive the fitting 
are
put in the wire-based track list and serve as the input to the 
time-based tracking phase.    Each fitted track in the time-based phase 
is then matched to the outer detectors (TOF and BCAL) and the dEdx in 
the chambers is calculated.   A figure-of-merit using the chi2 of the 
fit, the dEdx, and the time-of-flight to the outer detectors is computed 
to indicate how good the particular mass hypothesis (passed on through 
the wire-based fits) is.  I have created a new factory in the PID 
library called DChargedTrack that is intended to be the charged-particle 
equivalent of DPhoton.   The DChargedTrack object contains a vector of 
fitted tracks.  There is one vector for each track candidate.  The lists 
are ordered according to the figure-of-merit from the DTrackTimeBased 
object.   The DParticle factory is now reserved for a higher level of 
the reconstruction chain and is currently an empty shell.

Please let me know if you have any problems stemming from the 
reorganization.

Simon