Brunel-like options

Purpose

Update of hlt2_reco_brunelesque after comparing counters to the Brunel implementation. The goal, to get exactly the same output is reached with this MR. This helps to understand potential differences in the reconstruction that builds up on the track reco.

Current status

One thing to address when refactoring the changes is this warning:

UserWarning: multiple matches: higher-level True (standalone.py:90) shadows False (hlt2_tracking.py:215)
  return TrackBestTrackCreator(TracksInContainers=track_list, Fitter=get_fitter_tool(), GhostIdTool=get_ghost_tool(), InitTrackStates=False, DoNotRefit=False, AddGhostProb=True, FitTracks=True).TracksOutContainer

Resolved now by explicitly passing do_not_refit as argument to the maker (which is also not really nice)

Removed WIP after running local tests. It can now be tested in the nightlies. I expect the tests with references to fail.

Code changes

RecoConf.hlt1_tracking

• Made cut of global event cut an argument, since passing it as **kwargs resulted in conflicting arguments when passing it to PrGECFilter

RecoConf.hlt2_tracking

• Added two non-configurable functions get_default_hlt2_tracks and get_default_order_of_track_list_for_TrackBestTrackCreator which enable the configuration of the list of tracks consumed by TrackBestTrackCreator

RecoConf.standalone

• Applies the changes described above to set up the track reconstruction in the same way as Brunel