Refactoring of the caloriemter base code and bug fixes

[bechenard]

Major (and hopefully final) refactoring of the calorimeter code, including a few bug fixes, to facilitate condition integration and improve readability
o Changed interface to navigation classes. There is still a method to call mu2e<->tracker frame transformations (a big wart) that needs to be moved to the tracker and removed
o The SimParticle selection for the CaloShowerStep reconstruction was slightly tightened -> doesn't change the physics, but it changes the RNG sequence and the validation plots
o Kept the calorimeter interface to avoid exposing some of the DiskCalorimeter methods -> fixed some call sites

Will fix the CMakeList files when the tests complain...

[FNALbuild]

Hi @bechenard,
You have proposed changes to files in these packages:

• Analyses
• CalPatRec
• Mu2eG4
• EventGenerator
• EventDisplay
• Mu2eKinKal
• TrkPatRec
• CaloReco
• CaloDiag
• Filters
• CalorimeterGeom
• CaloMC
• CaloCluster
• TrkReco
• CaloFilters
• GeometryService

which require these tests: build.

@Mu2e/fnalbuild-users, @Mu2e/write have access to CI actions on main.

⌛ The following tests have been triggered for 521583d: build (Build queue - API unavailable)

About FNALbuild. Code review on Mu2e/Offline.

[FNALbuild]

☔ The build tests failed for 521583d.

Test	Result	Details
test with	✅	Command did not list any other PRs to include
merge	✅	Merged 521583d at 2818851
build (prof)	✅	Log file. Build time: 04 min 27 sec
ceSimReco	✅	Log file.
g4test_03MT	✅	Log file.
transportOnly	✅	Log file.
POT	✅	Log file.
g4study	✅	Log file.
cosmicSimReco	✅	Log file.
cosmicOffSpill	✅	Log file.
ceSteps	✅	Log file.
ceDigi	✅	Log file.
muDauSteps	✅	Log file.
ceMix	✅	Log file.
rootOverlaps	✅	Log file.
g4surfaceCheck	❌	Log file. Return Code 0.
trigger	❌	Log file. Return Code 2.
check_cmake	❌	Log file. Return Code 1.
FIXME, TODO	➡️	TODO (19) FIXME (38) in 61 files
clang-tidy	➡️	76 errors 3304 warnings
whitespace check	✅	no whitespace errors found

N.B. These results were obtained from a build of this Pull Request at 521583d after being merged into the base branch at 2818851.

For more information, please check the job page here.
Build artifacts are deleted after 5 days. If this is not desired, select Keep this build forever on the job page.

[bechenard]

@FNALbuild run build test

[FNALbuild]

⌛ The following tests have been triggered for 3cba7b1: build (Build queue - API unavailable)

[FNALbuild]

☀️ The build tests passed at 3cba7b1.

Test	Result	Details
test with	✅	Command did not list any other PRs to include
merge	✅	Merged 3cba7b1 at 2818851
build (prof)	✅	Log file. Build time: 04 min 24 sec
ceSimReco	✅	Log file.
g4test_03MT	✅	Log file.
transportOnly	✅	Log file.
POT	✅	Log file.
g4study	✅	Log file.
cosmicSimReco	✅	Log file.
cosmicOffSpill	✅	Log file.
ceSteps	✅	Log file.
ceDigi	✅	Log file.
muDauSteps	✅	Log file.
ceMix	✅	Log file.
rootOverlaps	✅	Log file.
g4surfaceCheck	✅	Log file.
trigger	❌	Log file. Return Code 2.
check_cmake	✅	Log file.
FIXME, TODO	➡️	TODO (19) FIXME (38) in 61 files
clang-tidy	➡️	76 errors 3304 warnings
whitespace check	✅	no whitespace errors found

N.B. These results were obtained from a build of this Pull Request at 3cba7b1 after being merged into the base branch at 2818851.

For more information, please check the job page here.
Build artifacts are deleted after 5 days. If this is not desired, select Keep this build forever on the job page.

[bechenard]

Trigger test invalidated because the RNG sequence is altered by changes. The rest should be ok

[oksuzian]

1. [S0] Crystal neighbor cache is now disk-LOCAL but consumed as GLOBAL → silent mis-clustering on disk 1. In GeometryService/src/DiskCalorimeterMaker.cc:259-261:
   //precompute the neighbors in the global frame // stale comment
   thisCrystal.setID(icry+crystalOffset); // id GLOBAL
   thisCrystal.setNeighbors(thisDisk.neighbors(icry,1)); // neighbors LOCAL (no offset)
   Disk::neighbors returns local ids; base used neighborsByLevel(icry+crystalOffset) (global), and DiskCalorimeter::neighbors(globalId,level) still adds the offset. So the cache now holds local ids where global are expected. For a disk-1 seed (globals 674–1347, 674/disk × 2 disks) crystal(id).neighbors() returns 0–673 → resolves to disk-0 crystals. It's silent (ids stay in range) and disk-0 works, so single-disk smokes pass. Affected: ClusterFinder (→CaloClusterMaker), CaloClusterFast, CaloTrigger (this PR changed it from the correct cal->neighbors(id) to the cache), ClusterUtils. Fix: offset in the maker before caching (v += crystalOffset), or switch consumers to cal->neighbors(globalId,level). Please add a disk-1 clustering check — this bug class is invisible to disk-0-only tests.

2. [S1] crystalPos.txt is loaded at runtime but not installed by CMake. Disk::fillCrystals() opens it via ConfigFileLookupPolicy (geom key diskCrystalFile), but CalorimeterGeom/CMakeLists.txt has no install(DIRECTORY data …) (sibling packages
   Mu2eG4/TrackerConditions/ParticleID all do). A CMake/spack install or Phase-2 tarball ships without it → DISK_OPEN_FAILED. The scons buildtest reads from the source tree so CI is green regardless. One-line fix: install(DIRECTORY data DESTINATION
   ${CMAKE_INSTALL_DATAROOTDIR}/Offline/CalorimeterGeom).

3. [S2] DiskCalorimeter copy/move ctors drop trackerCenter_ (DiskCalorimeter.cc:22-36) → zeroed on any copy → mu2eToTracker/trackerToMu2e become the identity (multi-mm error in track-calo matching). Latent (production path is a unique_ptr move) but a real regression; add trackerCenter_(rhs.trackerCenter_) to both.

4. [S2] The usePhysVolInfo FHiCL knob is now silently dead (CaloShowerStepMaker): the phys-vol branch was dropped but usePhysVol_/mapPhysVol_ remain (rebuilt every subrun, no reader). Setting it true is silently ignored. No production impact (default false); remove or re-wire.

5. [S3] isContainedSection→isInsideSameDisk is a real logic change (this is the RNG/plot diff) — energy-conserving, drops nothing at the production eDepThreshold: 0; add a comment. caloMotherXorig is stored sign-flipped vs its comment (dead key). Minor CMake/SConscript lib drift; dead Disk::radiusInCrystal_; stale "global frame" comment.

[oksuzian]

@FNALbuild run build test

[FNALbuild]

⌛ The following tests have been triggered for 35d9c05: build (Build queue - API unavailable)

[FNALbuild]

📝 The HEAD of main has changed to f96d9fc. Tests are now out of date.

[FNALbuild]

☀️ The build tests passed at 35d9c05.

Test	Result	Details
test with	✅	Command did not list any other PRs to include
merge	✅	Merged 35d9c05 at 2818851
build (prof)	✅	Log file. Build time: 04 min 21 sec
ceSimReco	✅	Log file.
g4test_03MT	✅	Log file.
transportOnly	✅	Log file.
POT	✅	Log file.
g4study	✅	Log file.
cosmicSimReco	✅	Log file.
cosmicOffSpill	✅	Log file.
ceSteps	✅	Log file.
ceDigi	✅	Log file.
muDauSteps	✅	Log file.
ceMix	✅	Log file.
rootOverlaps	✅	Log file.
g4surfaceCheck	✅	Log file.
trigger	✅	Log file.
check_cmake	✅	Log file.
FIXME, TODO	➡️	TODO (19) FIXME (38) in 61 files
clang-tidy	➡️	76 errors 3305 warnings
whitespace check	✅	no whitespace errors found

N.B. These results were obtained from a build of this Pull Request at 35d9c05 after being merged into the base branch at 2818851.

For more information, please check the job page here.
Build artifacts are deleted after 5 days. If this is not desired, select Keep this build forever on the job page.

[giro94]

Ok for me. This might have some consequence in EventNtuple code, e.g. https://github.com/Mu2e/EventNtuple/blob/main/src/InfoStructHelper.cc#L532C1-L532C79

[oksuzian]

Findings

1. [S1] crystalPos.txt is still not installed by the CMake build — and the new configure_file line doesn't fix it.

The file is loaded at runtime via ConfigFileLookupPolicy (geom key calorimeter.diskCrystalFile, throws DISK_OPEN_FAILED if absent). The new line in CalorimeterGeom/CMakeLists.txt:

configure_file(${CMAKE_CURRENT_SOURCE_DIR}/data/crystalPos.txt ${CURRENT_BINARY_DIR} data/crystalPos.txt)
has two problems:
- ${CURRENT_BINARY_DIR} is not a defined CMake variable (the standard one is CMAKE_CURRENT_BINARY_DIR); it expands to empty, which silently collapses the call to a 2-arg configure_file that copies the file into the build tree.
- Even working as intended, configure_file is a configure-time copy, not an install rule — a CMake/spack install or a Phase-2 grid tarball still ships without the file, and the first calo-geometry job dies. The scons Jenkins CI can't catch this
(it reads the source tree).

Fix (one line, matches every sibling package that ships runtime data — Mu2eG4/CMakeLists.txt:306-310, TrackerConditions, ParticleID):
install(DIRECTORY data DESTINATION ${CMAKE_INSTALL_DATAROOTDIR}/Offline/CalorimeterGeom)
and drop the configure_file line.

2. [S3] mapPhysVol_ is now a write-only set rebuilt every subrun.

With the usePhysVol_ branch gone, CaloShowerStepMaker_module.cc still clears and repopulates mapPhysVol_ in beginSubRun (lines ~195-207), but nothing reads it anymore. Remove the member and the population loop (physVolInfoInput_/vols_ are still
genuinely needed by PhysicalVolumeMultiHelper).

3. [S3] Minor / confirm (carried over from the first pass, unchanged)
- isContainedSection → isInsideSameDisk (CaloShowerStepMaker): a genuine logic change (old = Δz-only; new = both endpoints inside the disk envelope), not a rename — this is the RNG/validation-plot change the author flagged. It conserves
per-crystal deposited energy at the production eDepThreshold: 0. A one-line comment that the criterion changed deliberately would help future archaeology.
- The shim-step loop assumes every crystal row straddles x=0 (sentinels xLowMin=1e6…); fine for the shipped crystalPos.txt (all 37 rows have both sides), fragile for future asymmetric files.
- CMake vs SConscript link-lib drift (cetlib on scons only); cosmetic. Dead Disk::radiusInCrystal_ member. DiskCal00_module.cc dropped an Origin: debug print rather than porting it. SquareShiftMapper::colFromIndex changed the negative-column
convention (now self-consistent, internally unused — confirm no external caller relied on the old numbering).