Support extras (crack, topgroove, bottom_cylinder, top_cylinder) in SSD config file creation

[giriPHM]

… Refactored and updated from @apearsonn's work in PR #81. Integrated logic into the modern extension structure and verified compatibility with the current metadata pipeline. I did run the tests as well. All the tests passed. There was only one warning that the test seemed to generate. I have attached the screenshot for the output from the test below.

[codecov]

Codecov Report

❌ Patch coverage is 90.24390% with 16 lines in your changes missing coverage. Please review.
✅ Project coverage is 50.49%. Comparing base (d2134f7) to head (cfab173).

Files with missing lines	Patch %	Lines
ext/LegendDataManagementSolidStateDetectorsExt.jl	89.80%	16 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #157      +/-   ##
==========================================
+ Coverage   47.85%   50.49%   +2.64%     
==========================================
  Files          27       27              
  Lines        2466     2533      +67     
==========================================
+ Hits         1180     1279      +99     
+ Misses       1286     1254      -32     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[hervasa2]

Thanks for the initiative! There are many changes that undo recent updates which have nothing to do with the additional geometry support. Please undo these and address the ignoring the Li thickness comment. My recommendation is use something with exaggerated n_thickness and plot to check like so:

sim = Simulation{T}(LegendData, meta, xtal_meta, verbose = true, HPGeEnvironment("LAr", 87u"K"), n_thickness = 5u"mm");
plot(sim.detector, st = :slice)

Could you also please include a test of each new feature, showing a snippet of the metadata and the resulting plot (using exaggerated n_thickness). Also test one detector with no "extras" to see that the old functionality still works.

[Copilot]

Pull request overview

Adds support in the SolidStateDetectors extension for additional LEGEND SSD geometry “extras” (bottom cylinder, top groove, crack) when generating SSD configuration dictionaries from metadata, refactoring logic previously explored in PR #81 to fit the current extension/metadata pipeline.

Changes:

• Add handling for geometry.extra.bottom_cylinder and geometry.extra.topgroove in the generated semiconductor/contact geometries.
• Add handling for geometry.extra.crack, including forcing 3D simulation (phi to 360) and cutting the crack volume out of semiconductor and mantle contact geometry.
• Update phi axis boundary behavior to "periodic" to support 3D cases.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[fhagemann]

This should also be extended to include support for top_cylinder for V06649M (see #159 and https://github.com/legend-exp/legend-detectors/pull/74)

[giriPHM]

This should also be extended to include support for top_cylinder for V06649M (see #159 and legend-exp/legend-detectors#74)

Thanks for pointing that out

[giriPHM]

I have used

sim_2 = Simulation{T}(LegendData, meta, xtal_meta, verbose = true, HPGeEnvironment("LAr", 87u"K"), n_thickness = 2u"mm");
plot(sim_2.detector, st = :slice, legend=:topright, title="$det")

The top taper for V02160A and V02162B might look a bit weird, but that is just because the angle is small and the height of the taper is scaled by tangent of the angle.

[giriPHM]

Since the n+ contact geometry of V02162B (one with the top groove) looked a bit weird, I went back and made some fixes to the geometry. Now, I think it looks good. Since the top groove height is just 2.7 mm, I have used n_thickness of 1.5 mm to see things better.

[theHenks]

@giriPHM Thank you for your work here. Are you certain that all this metadata is public? Since we are sharing this in a public repository, I recommend double-checking what can be shown and what cannot to ensure we do not release any confidential detector data.

[giriPHM]

@giriPHM Thank you for your work here. Are you certain that all this metadata is public? Since we are sharing this in a public repository, I recommend double-checking what can be shown and what cannot to ensure we do not release any confidential detector data.

That is a good point, Florian. I was posting those snippets in response to David's comment regarding the test of the features. But I will remove that information till we figure this issue out!

[giriPHM]

@Copilot. Can you suggest some tests for the added features to address the patch coverage?

[giriPHM]

Add top cylinder feature with tests; fix borehole depth parameter in active_volume.jl

• Implemented top cylinder as a new geometry feature
• Added corresponding tests for extra features functionality
• Fixed bug in active_volume.jl where borehole depth was incorrectly using radius parameter

outdated

[hervasa2]

• Refactored mantle_contact implementation to include extras and remove redundancies. The result is much more compact and maintainable.

• Taper mantles are now handled correctly with a dedicated function.

• Added tests for each extra. This involves creating one base detector and removing the different extras from the base volume. A point from the removed volume is tested to be in the original detector but outside of the modified volume. An additional test is added to check that the active volume of the modified volume is smaller than the base detector active volume.

• Base:

• Top cylinder:

• Bottom cylinder:

• Top groove:

• Crack:

[fhagemann]

In order to review this conceptually (without looking at every line of code): can you also attach picture for the „normal“ mantle taper case. Just to make sure that the refactoring did not break anything? :)

[hervasa2]

See exaggerated "normal" mantle taper cases without extras:

• IC

• COAX

• PPC

• BEGe

[fhagemann]

Looks good! One final comment

[fhagemann]

I guess someone (@theHenks / @oschulz ) can merge this now :)