[Halld-offline] Newest mcsmear change

[beattite]

Hello.

I just checked in some changes before the svn to Git switch.

First, an updated shower curvature code is up 
(DBCALShower_factory_CURVATURE.cc) which just implemented a more 
variable (with shower theta) z-bin width.  It's gives a minor 
improvement in point acceptance (Revision 19022).

Second, the transition from mcsmear outputting Hits to DigiHits.  Six 
files were changed here.
  DEventSourceHDDM.cc and DEventSourceHDDM.h - Removed filling of 
DBCALHits from HDDM structure and added filling of DBCALDigiHits.
  event.xml - Added declaration of DigiHits data structure (using 
pulse_integral and pulse_time as integers rather than E and t as 
floats).
  mcsmear.cc and smear_bcal.cc - Converted final hit energy and time to 
ADC counts using some known constants (0.065 ns per ADC count for time, 
0.029 MeV per integrated ADC count for energy) before outputting to HDDM 
DigiHits structure rather than the Hits structure.
  DBCALHit_factory.cc - Changed this a bit so it handles simulated 
DigiHits properly.  It'll set the pedestal to zero and 
n_samples_pulse/n_samples_pedestal to appropriate values so the total 
integral calculation reports the correct numbers.

The 'variation=mc' ADC_gains CCDB table exists, as is populated with 
values of 1e-4.  Using this, the new code gives a reconstructed cluster 
energy of around 3.3 GeV for a 1 GeV photon gun at 90 degrees, so when 
we do the calibration, the gains should come out to around 3e-5.

This second revision was Revision 19023.