[Newton] Streamline contact and raycast sensor startup

[camevor]

Summary

This PR is now the contact/raycast part of the Newton startup work:

• compile contact-sensor full-path expressions once and match them directly, without regex-to-glob conversion or stage globbing;
• use Newton's current contact-sensing API names;
• declare raycast collision-shape requirements before model finalization, so the model builds one correctly configured BVH instead of rebuilding it during sensor initialization.

The contact-selector and sensing-API commits retain Chris's (camevor) original authorship. The BVH lifecycle change incorporates the review from #7296 while keeping this as Chris's PR.

The related ownership work is intentionally split by component: #7292 handles explicit global ownership, and #7295 handles model/articulation startup. This PR contains no inactive-solver registration or articulation-view changes and supersedes #7296.

Production code is +29/-56 (net -27) against 909cc5decc5.

Startup benchmark

RTX 5090, CUDA device 1, 4096 environments, three fresh processes per revision/task. Values are median end-to-end startup wall time. Base: 909cc5decc5. PR: 5d0e1b1595e.

Task	Base	PR	Change
Isaac-Cartpole	8.316 s	8.410 s	+1.1%
Isaac-Velocity-Rough-UnitreeGo2	16.363 s	15.682 s	-4.2%
Isaac-Lift-KukaAllegro-Camera	39.473 s	37.605 s	-4.7%

Cartpole has no contact/raycast workload here and is neutral within process-startup noise. The sensor-heavy tasks show the intended gain:

Median phase	Go2 base	Go2 PR	Kuka base	Kuka PR
newton_contact_sensor	0.10 s	0.04 s	1.32 s	0.24 s
simulation_start	6.36 s	5.91 s	11.48 s	10.39 s

Raw end-to-end totals:

• Cartpole base: 9.368, 8.316, 8.195 s; PR: 8.410, 8.370, 8.476 s.
• Go2 rough base: 19.900, 16.363, 16.091 s; PR: 15.669, 15.682, 15.984 s.
• Kuka camera base: 41.328, 38.625, 39.473 s; PR: 37.580, 37.605, 38.785 s.

Test plan

• 261 passed, 8 xpassed across the Newton manager abstraction, contact-sensor, and raycast-sensor suites.
• Repository formatting and pre-commit checks pass.
• Architecture checks reject the removed regex-to-glob path, deprecated sensing names, duplicate BVH state, and late BVH fallback.
• The three 4096-environment benchmark tasks also provide end-to-end Newton MJWarp startup coverage.

Type of change

• Performance improvement
• Bug fix

[greptile-apps]

Greptile Summary

The PR reduces Newton startup work while correcting path-expression handling and preserving solver-specific USD attributes.

• Registers custom builder attributes only for active direct and coupled solvers.
• Resolves asset and contact-sensor selectors as full-match regular expressions.
• Includes collision shapes in the finalized BVH when a ray caster requests them.
• Migrates contact-sensor integration to Newton’s current sensing API and adds regression coverage.

Confidence Score: 5/5

The PR appears safe to merge, with no concrete blocking or independently actionable non-blocking issue identified.

The changed builder, selector, and BVH lifecycles retain direct-stage, cloned, coupled, and post-finalization fallback paths, and the accompanying tests cover the principal behavior changes.

Important Files Changed

Filename	Overview
source/isaaclab_newton/isaaclab_newton/physics/newton_manager.py	Centralizes regex compilation, active-solver attribute registration, and pre-finalization sensor BVH requirements without an accepted correctness issue.
source/isaaclab_newton/isaaclab_newton/sensors/contact_sensor/contact_sensor.py	Migrates to Newton’s current sensing fields and retains initialization and buffer-shape validation.
source/isaaclab_newton/isaaclab_newton/sensors/ray_caster/newton_raycast_sensor.py	Requests collision-shape BVH coverage before finalization while retaining the post-finalization rebuild fallback.
source/isaaclab_newton/isaaclab_newton/assets/articulation/articulation.py	Passes articulation-root expressions directly as regular expressions, avoiding lossy glob conversion.
source/isaaclab_contrib/isaaclab_contrib/coupling/coupler.py	Registers nested MJWarp attributes when needed while preserving existing MPM handling.
source/isaaclab_newton/isaaclab_newton/cloner/newton_clone_utils.py	Removes unconditional solver imports and relies on the active manager’s pre-import builder factory.

Sequence Diagram

sequenceDiagram
    participant Scene as Scene construction
    participant Sensor as Newton ray/contact sensor
    participant Manager as Active Newton manager
    participant Builder as ModelBuilder
    participant Model as Finalized model
    Scene->>Sensor: Instantiate configured sensors
    Sensor->>Manager: Request contact data or BVH shape flags
    Scene->>Manager: Create builder using active solver manager
    Manager->>Builder: Register active solver attributes
    Manager->>Builder: Import or replicate USD model
    Manager->>Builder: Apply pending BVH requirements
    Builder->>Model: Finalize once
    Manager->>Sensor: Dispatch physics-ready initialization
    Sensor->>Model: Bind regex-selected views and sensor buffers

Loading

_{Reviews (1): Last reviewed commit: "fix test" | Re-trigger Greptile}

[isaaclab-review-bot]

Isaac Lab Review Bot

Reviewed the Newton startup-time changes across solver-specific builder attribute registration, pre-finalization ray-caster BVH configuration, and compiled full-path regex selectors for contact sensors and Newton views. No numbered candidate findings were supplied for acceptance.

• Design and architecture: The active manager now owns solver-specific attribute registration, including nested coupled solvers, while ray casters request collision-shape BVH coverage before model finalization and retain the existing late-registration rebuild fallback.
• API: Checked the compatibility surface for contact-sensor shape selectors, body selectors, and Newton view roots. The intentional full-path, full-match regex behavior is documented with migration examples in configuration docstrings and package changelog fragments; public configuration field names remain unchanged.
• Implementation: Traced custom-attribute registration through builder creation, stage import, coupled managers, and finalization. Also traced BVH requirements from ray-caster construction through finalization, fallback rebuilding, and manager cleanup, plus the selector path from expression compilation to Newton sensor/view construction. The added tests cover solver-registration ownership, BVH inclusion and fallback behavior, regex matching, malformed expressions, and coupled-manager registration.

No blocking issues. No inline issue met the actionable-evidence threshold; the assessment above records the review feedback.

Automated review; human maintainers own approval decisions.