ParticleHandler: account for modified stepping actions, where tracking…

[andresailer]

… of a particle could be paused and later restarted.

Pick up the particle information that we have stored and go on from there

BEGINRELEASENOTES

• ParticleHandler: account for modified stepping actions, where tracking of a particle could be paused and later restarted.

ENDRELEASENOTES

For #1036 I get this for 2 events now

root [2] EVENT->Scan("sqrt(MCParticles.psx^2 + MCParticles.psy^2 + MCParticles.psz^2 + MCParticles.mass^2):MCParticles.reason:MCParticles.pdgID:MCParticles.mask", "MCParticles.mask==1")
***********************************************************************
*    Row   * Instance * sqrt(MCPa * MCParticl * MCParticl * MCParticl *
***********************************************************************
*        0 *        0 *     10000 *       158 *        11 *         1 *
*        1 *        0 *     10000 *       158 *        11 *         1 *
***********************************************************************
==> 2 selected entries
(long long) 2

So the energy is the same as the initial one (10000) as opposed to some smaller number.

[andresailer]

@MarkusFrankATcernch Does this look sensible to you?

[MarkusFrankATcernch]

If the tracking action is the first call to the particle handler after the restacking, this should be correct.
However, I have the suspicion that all properties of the existing "currentTrack" may all somehow lost, because this
track and its properties is not yet saved away to the collection possibly be kept in the output record.

Hence the question is: Whan does the restacking happen? before or after a couple of steps were already processed?

[MarkusFrankATcernch]

if before the restacking the member

void Geant4ParticleHandler::end(const G4Track* track)

is called this approach should be correct. Be aware that this track may be in both collections:

• Geant4ParticleHandler::m_equivalentTracks and
• Geant4ParticleHandler::m_particleMap

in Geant4ParticleHandler::endEvent(const G4Event* event) preference must be given to Geant4ParticleHandler::m_particleMap when rebasing the tracks.

[andresailer]

I implemented the treatment also for particles that might not be in the particleMap before they are suspended. It looks fine to me, but that code is never actually run so far.

[andresailer]

I am adding a test

[andresailer]

Added a test that exercises the suspendedPM related code