[Halld-offline] Tracking code reorganization

[Simon Taylor]

Hi, everyone.

Based on discussions with Dave I have reorganized the tracking code to 
make it more transparent what is going on:  the DTrack object in the 
TRACKING library has been renamed DTrackWireBased because it represents 
the results of a wires-only fit to the data.   The results of  the 
time-based fit are stored in the new DTrackTimeBased object, whose 
factory is a copy of the old DParticle factory in the PID library.   I 
propose to reuse the DParticle factory as a true particle factory with 
particle id including information from the outer detectors.  I have
checked in numerous small changes to the PID factory in addition to 
several larger changes to the DTrack factory.   In particular those of 
you who are using the DPhoton factory will need do an update to remain 
in sync with the tracking library (although it looks like the track 
matching using the track fit results is disabled right now).   This is a 
temporary solution for the DPhoton_factory -- eventually the code 
explicitly using  DMCThrown data should be removed.   I propose that the 
functionality of the DPhoton factory be merged into the DParticle 
factory where track matching and particle id for all particles can take 
place and we end up with a list of both charged and neutral particles.

Please let me know if you experience any problems.

Simon