Chrysler: match firmware revisions with required control ECU evidence

[khchen428]

Replaces #3733 to submit this work from khchen428. The code and commits are unchanged.

The validation described below applies to these unchanged commits and includes results from the original PR. CI for this replacement PR is pending.

Recognize unseen Chrysler firmware revisions by matching the complete eight-character part number while ignoring its two-letter software revision. Require the candidate's known EPS, ABS, radar, and SRS parts, preserve exact matching for opaque radar responses, and tolerate one missing or previously unknown cluster part when other expected modules match. Known conflicting cluster parts and ambiguous candidates remain rejected.

Add a default-on use_generic_fuzzy option and disable it for Chrysler: otherwise generic matching can accept a vehicle from other ECUs before the custom matcher checks unknown or missing control-module firmware. A public-entry regression test demonstrates this bypass and verifies the opt-out. Existing brands retain their current behavior.

Validation:

• Fifteen Chrysler tests covering every checked-in firmware version, 2,000 deterministic known/revised combinations, all 90 directed platform pairs, malformed responses, missing/conflicting modules, and public matcher integration.
• Full ./test.sh passes: 3,986 tests run, 3,294 passed and 692 skipped; Ruff, ty, codespell, cpplint, and MISRA pass.
• A pinned commaCarSegments sample supplies 64 firmware snapshots from 45 devices. Both baseline and proposed matching agree with 63/64 recorded labels, with no wrong or ambiguous matches. The remaining Durango has new control parts and is rejected. Two RAM HD devices with new cluster parts and known control parts motivate the cluster exception.
• With all parseable software revisions synthetically replaced by ZZ, proposed matching preserves those 63 matches; baseline matches none.
• An actual historical holdout against the December 2024 database recognizes an unseen EPS revision by part number, but deliberately rejects four additional snapshots with genuinely new ABS/EPS parts: 59/64 matches versus generic matching's 63/64. This is a documented recall cost of requiring known control parts. Logged labels are not independent compatibility evidence.

Dataset revision: edb6480d0e0b5775fde7150c01bcdd5b78d0ad32; historical database: fe27b134350c44724aa2078de12d1c447f8e3083. No hardware was used.

Related to #1092. New control-module part numbers still require validation and database updates; this does not claim to identify every future compatible model.

[github-actions]

Thanks for contributing to opendbc! In order for us to review your PR as quickly as possible, check the following:

• Convert your PR to a draft unless it's ready to review
• Read the contributing docs
• Before marking as "ready for review", ensure:
  • the goal is clearly stated in the description
  • all the tests are passing
  • include a route or your device' dongle ID if relevant

[github-actions]

Car behavior report

Replays driving segments through this PR and compares the behavior to master.
Please review any changes carefully to ensure they are expected.

Testing 40 segments for: CHRYSLER_PACIFICA_2018_HYBRID, CHRYSLER_PACIFICA_2019_HYBRID, CHRYSLER_PACIFICA_2018, CHRYSLER_PACIFICA_2020

✅ 0 changed, 40 passed, 0 errors