ddsim: support non-cylindrical tracking region

[Zehvogel]

I would like to have a non-cylindrical tracking region as introduced in #384, particularly because in the naive cylindrical approach the lumi cal of my detector is inside the tracking volume.

When I add such a tracking region to my detector and remove the previous definition of the tracking region:

        <constant name="tracker_region_zmax" value="OuterTracker_half_length"/>
        <constant name="tracker_region_rmax" value="OuterTracker_outer_radius"/>

ddsim stops here:

DD4hep/DDG4/python/DDSim/Helper/ParticleHandler.py

Lines 129 to 139 in 7967bdf

	if self.userParticleHandler == "Geant4TCUserParticleHandler":
	user = DDG4.Action(kernel, "%s/UserParticleHandler" % self.userParticleHandler)
	try:
	user.TrackingVolume_Zmax = DDG4.tracker_region_zmax
	user.TrackingVolume_Rmax = DDG4.tracker_region_rmax
	except AttributeError as e:
	logger.error("Attribute of tracker region missing in detector model %s", e)
	logger.error(" make sure to specify the global constants tracker_region_zmax and tracker_region_rmax ")
	logger.error(" this is needed for the MC-truth link of created sim-hits ! ")
	logger.error(" Or Disable the User Particle Handler with --part.userParticleHandler=''")
	exit(1)

(If I don't remove it the cylinder is still used as the tracking region as far as I can tell.)

The TrackingRegion example in the examples/ClientTests seems to not use ddsim:
https://github.com/AIDASoft/DD4hep/blob/master/examples/ClientTests/scripts/TrackingRegion.py

[andresailer]

I think one would need to implement a Geant4TVUserParticleHandler, where TV is for tracking_volume? (Or TR for tracking region?). Maybe I missed the use of the tracking_volume?

[Zehvogel]

Aaah ok, the TC in Geant4TCUserParticleHandler stands for tracking cylinder... so there is no way yet to use this feature?

[andresailer]

As far as I saw there isn't one yet.
Do you want to give it a go to implement the new handler?

[Zehvogel]

In principle that looks simple. I just don't know how I would retrieve that volume in ParticleHandler.py to pass it to the handler... and I don't really know where to look to figure that out in this interesting c++ and python mixture :(

[Zehvogel]

nvm. I think I found a way :) I will prepare a PR