feat(geometry): tessellate IfcFaceBasedSurfaceModel

[Lutherwaves]

tessellateItemDepth had no case for IfcFaceBasedSurfaceModel, so every
element built from one degraded to a bounding box. It was the top entry of the
parity harness's measured gap list at 65 occurrences, and in duplex_a it hid
235 IfcConnectedFaceSets that no other path could reach.

Measured result

Model	OBB rate before	after
ifcopenhouse	0.0%	0.0%
duplex_a	32.1%	1.9%
fzk_haus	2.4%	2.4%

duplex_a's row moves 146 Extrude + 0 Brep + 69 OBB + 3 Empty to
146 + 65 + 4 + 3. Exactly 65 elements moved from OBB to Brep, Extrude and
Empty did not move at all, and the IFCFACEBASEDSURFACEMODEL row left the
unhandled table. The model had zero brep elements before this.

This is far past the ~13% the design doc predicted, so it is worth saying
why. That prediction came from an estimate that face-based surface models
accounted for ~58% of duplex_a's fallbacks — an estimate that conflated item
OCCURRENCES with fallback ELEMENTS, which is exactly the error
docs/coverage.md's own prose warns against. The measurement is the answer and
the estimate was the guess. The 65-occurrences/65-elements agreement is not a
coincidence either: occurrences are counted once per element that reaches an
item, so for a model with one such item per element the two figures are equal
by construction.

The residual 1.9% is four elements, and the only remaining entry in the
unhandled table is IFCPOLYGONALBOUNDEDHALFSPACE (11) — issue #52, the next PR.

Gate 1 is unchanged

Checked before regenerating anything, which is the order that matters:
make parity-baseline rewrites the file Gate 2 compares against, so running it
first would turn a failing gate green whether or not the change was correct.

331 elements compared, 329 contained, with exactly the two pre-existing
IfcStairFlight violations (1oKjKg9PD3fP1iIwXLh3lK min-y,
3KMJUyUe9DfQ2FOCd5ZoiN max-y) still present and still violating. No new
violation appeared, and parity/knownviolations.go needed no change. That
matters because this PR replaces conservative superset boxes with tight
meshes — the change most likely to expose an element that passed containment
only because its box was loose.

Shape of the change

FbsmFaces is attribute 0, a SET of IfcConnectedFaceSet. SbsmBoundary is
attribute 0, a SET of IfcShell. Different schema entities, identical
traversal, and brepMesh already tessellated a bare IfcConnectedFaceSet. So
the first commit generalises the existing shell-based helper into
surfaceModelMesh(m, attr) as a pure refactor, and the second adds the
dispatch case using it. The attribute is a parameter rather than assumed
because the two constants agreeing at 0 is a fact about the schema, not a rule.

Split that way so a bisect lands on the dispatch case rather than on the rename.

A note for whoever next touches brepMesh

surfaceModelMesh and brepMesh do not type-check set members — brepMesh
reads attribute 0 of whatever it is handed. The decline test documents this
honestly rather than claiming more than it checks: an IfcCartesianPoint gets
as far as its Coordinates list and declines only because that list holds
reals rather than references. A wrong-typed member whose attribute 0 were a
list of references would still be walked as a shell. Closing that would move
measured numbers, so it needs its own change with its own measurement pass.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

Next included review available in 48 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: 0d1ccd42-d9af-490a-a023-555c2bc01e34

📥 Commits

Reviewing files that changed from the base of the PR and between 551ac66 and 7dd759c.

📒 Files selected for processing (7)

• docs/coverage.md
• geometry/brep.go
• geometry/brep_test.go
• geometry/geometry.go
• geometry/testdata/synthetic/face_based_surface_model.ifc
• geometry/unhandled.go
• parity/testdata/baseline.json

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codecov]

Codecov Report

❌ Patch coverage is 66.66667% with 2 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
geometry/geometry.go	50.00%	2 Missing ⚠️

📢 Thoughts on this report? Let us know!