Make ConformalTracking to behave less diruptively on encountering too many tracks

[Zehvogel]

Before this, ConformalTracking would throw a SkipEventException resulting in further Marlin processors down the line to not be executed. However, when using Marlin processors from Gaudi with the k4MarlinWrapper, the expected behaviour of a failing processors is to (optionally warn and) emit its usual output collections. Otherwise PodioOutput crashes in the end...

BEGINRELEASENOTES

• Emit an error message and skip further steps when too many tracks are created instead of throwing a SkipEventException

ENDRELEASENOTES

I tested the changes locally and they solved key4hep/k4MarlinWrapper#158 for me. The track pulss also did not change more than they do from run to run anyway so I hope I did not break anything.

[andresailer]

Could you undo the unrelated whitespace changes?

Do you not have a global gitignore file?

Does this now potentially leave a partially reconstructed event or will no tracks be created?

[Zehvogel]

It will reconstruct as many tracks as far as it can. For the event in the linked issue it reconstructs 1 track (the correct one from the single pion) and results in a useful PFO with correct energy further down the chain.

[andresailer]

So it might give some badly reconstructed event, instead of no reconstructed object. But if needed one should be able to filter this out.

Thanks for for the changes!