Simplification of the 2D-strip re-segmentation of MPGDs

[ybedfer]

https://github.com/eic/EICrecon/blob/2D-strip-MPGDs/src/algorithms/digi/MPGDTrackerDigi.cc
(Note: the link supra points to the 2D-strip-MPGDs branch, before it's merged into main.)

The file implements a 2D-strip segmentation for MPGDs. It's a re-segmentation, overwriting what's done at simulation, as opposed to reconstruction, time and stored into the SimTrackerHit input to the reconstruction. 2D-strip means here a readout with a simultaneous registering of a unique ionization process along two distinct strip coordinates. The whole thing is achieved software-wise thanks to a 3-valued MultiSegmentation: 1 = pixel cellID (or pixelID) at simulation time, (2,3) = two strip cellIDs (or stripIDs) determined at reconstruction time by MPGDTrackerDigi.cc.

Describe the solution you'd like

• The determination of the two stripIDs is done by re-executing the segmentation process based on the Cartesian coordinates of the input SimTrackerHit via the dd4hep::Segmentation::cellID method.
• It turns out that the pixelID is the concatenation of the two stripIDs. In hexadecimal representation:
  pixelID = (hex1|hex2) , stripID1 = (hex1|0) , stripID2 = (0,hex2) .
  This, provided that the IDDescriptors of pixel and strips match.
• Therefore, we could simplify out dd4hep::Segmentation::cellID and simply split the input pixelID into two.
  The association RawTrackerHit -> SimTrackerHit could also be made simpler.

Additional context

• Keeping MPGDTrackerDigi.cc as is offers still some advantage: one can reconstruct successively a same input data set with different 2D-strip readouts, which can be useful in the debugging stage.