Full model based on GEN_10M.opc on UVA1

Before too long, we will need to modify g4sbs to be compatible with ROOT 6. The only aspect of the code causing issues right now is the "G4SBSRunData" class that stores the run metadata in the output ROOT file.

Radiative and multiple scattering effects needs to be included. Probably should be done with a volume vector that can be added to as components get added in DetectorConstruction

We would eventually like to make the g4sbs code capable of running in Multi-threaded mode. This is essential for the future and for large-scale simulations on the JLab batch farm. Given that the memory consumption of GEANT4 is so large, running g4sbs single-threaded is actually very inefficient even on the batch farm, because the farm nodes are configured at 32 GB RAM for 24 cores; i.e., 1.5 GB per core.

Field map file can be found here:

http://hallaweb.jlab.org/12GeV/SuperBigBite/downloads/GEN_10-map.txt

The axis Z is along the central ray to SBS (its front face is at z = 61 cm).
Target is located at x=0, y=0, z=341.
The beam line at 17.5 degrees
The BigBite at 51.5 degrees at 155 cm from target.
The corner of the table x=-250, y=-250, z=-200, steps are 2 cm.
Number of steps = 251, 251, and 301

If a user creates a macro that then executes another macro via the /control/execute command, the other macro does not get stored in the ROOT file.

Since the purpose is to keep track of the entire state in which G4SBS was called, then the contents of the external script should also be included in the ROOT file.

Right now we have a few scattered conditionals that check for some experiment names to decide what to build. But these are scattered across the different source files.

I think we should, instead, make flags that get configured in just one file. Ex:

if (kGEp) {
fBuild48D48 = true;
fBuildTargetSnout = true;
// ...
} else if (kGMn) {
fBuild48D48 = true;
fBuildTargetSnout = false;
// ...
}

That way, all the settings are in one file, and can be re-configured much quicker and hopefully, with less errors.

The INFN GEM simulation that provided rates from before provided output in a specific way so hits could be digitized. This code should have a separate tree which can be used with this.

This requires a full re-write of the generator system as well as incorporating the quasielastic information. A beam/target class like remoll also needs to be implemented (and set in the detector construction).