refactor(dc_measurements): deepen the Mission adapter lifecycle and pending-record queue

[Minipada]

Summary

Architecture-review-driven refactor of dc_measurements' Mission Measurement adapter family (nav2 NavigateToPose/NavigateThroughPoses/FollowWaypoints, Open-RMF) plus Fault/Ros2ControlStatus — the hottest area of the repo by recent commit volume. All three review candidates are implemented.

1. Mission adapter lifecycle boilerplate

The four Mission adapters converge on one ADR-0010 Record shape but each duplicated the outcome vocabulary, UUID formatting, JSON record building, and bounded-id bookkeeping. Extracted into small, single-purpose modules under dc_measurements/include/dc_measurements/:

• mission_outcome.hpp — one shared MissionOutcome enum + missionOutcomeName() + MissionStartFact, replacing four duplicated copies.
• mission_registry.hpp — PrunedMissionMap<Active>, the bounded id → state bookkeeping (oldest-finished-first eviction, sequence counter) MissionNav2ThroughPosesCore and MissionOpenRmfCore had each reimplemented identically.
• mission_record_json.hpp — shared missionStartJson()/missionEndJsonBase() builders, replacing MissionNav2/MissionNav2FollowWaypoints' static methods and the inline JSON blocks in the other two adapters.
• mission_uuid.hpp — the UUID formatter duplicated between MissionNav2 and MissionNav2FollowWaypoints. MissionNav2ThroughPoses' distinct (non-dashed) mission_id format is kept as a separate function rather than unified, since collapsing the two would silently change an already-shipped adapter's mission_id format.

Bug fix riding along: MissionNav2ThroughPoses' status subscription was on plain QoS(10) while its two siblings use reliable().transient_local() for late-joiner correctness. It now matches — this drift was direct evidence the duplication was letting real bugs through.

Deliberately not done: merging MissionNav2Tracker/MissionFollowWaypointsTracker's state-machine internals. Once the above is factored out, what differs between them is each endMission()'s extra parameter (recoveries vs missed_waypoints) — a small enough remainder that a shared base would add indirection for little gain.

2. Pending-record queue

pending_record_queue.hpp — the capped deque + drop-oldest + throttled-warning pattern duplicated across all four Mission adapters plus Fault and Ros2ControlStatus (six call sites).

3. Shared Mission Record JSON Schema

The four mission_*.json schemas duplicated ~90% of their property definitions — the schema-level mirror of the C++ duplication above. mission_base.json now holds the shared property definitions (event, mission_id, sequence, outcome, reason, error_code, duration_sec); each adapter schema allOf/$refs it and adds only its own mission_type/extra-property/required/conditional rules.

Cross-file $ref needs a loader callback that nlohmann_json_schema_validator's default json_validator doesn't have (every existing schema in this codebase kept to same-file #/$defs/... refs, which is why this was never needed before — confirmed against the real library: without a loader, set_root_schema()/validate() throws "external schema reference ... needs loading, but no loader callback given"). Added makeSchemaFileLoader() to measurement.hpp, wired into both Measurement::validateSchema() overloads. Four test files each independently re-validate emitted Records via their own locally-constructed json_validator as a self-check; each needed the same loader wired in.

Verification

Public class names, method signatures, and Record JSON shapes are unchanged (aside from the QoS fix above and the schema restructuring, which is validation-semantics-preserving — see below).

This was built and tested for real against a container image with the actual ROS 2 Jazzy toolchain (not just reviewed by hand):

• colcon build --packages-up-to dc_measurements: clean, zero errors, under -Wall -Wextra -Wpedantic -Werror -Wdeprecated.
• colcon test --packages-select dc_measurements: 293 tests, 0 errors, 0 failures, 0 skipped.
• The schema restructuring was additionally verified in isolation: 21 representative Records (valid and invalid) validated identically before/after the mission_base.json extraction, run directly against nlohmann_json_schema_validator.

Test plan

• colcon build (dc_measurements) — verified in CI-equivalent container
• colcon test — 293/293 passing, including all Mission/Fault/Ros2ControlStatus tests
• prek run --all-files (clang-format / lint) — not run locally (no clang-format binary in this environment); please run in CI

🤖 Generated with Claude Code

https://claude.ai/code/session_01VoPZP5DPoJuuCWtctZAfBe

[codecov]

Codecov Report

❌ Patch coverage is 91.02564% with 14 lines in your changes missing coverage. Please review.
✅ Project coverage is 70.60%. Comparing base (0ed5a64) to head (c148bc9).
⚠️ Report is 1 commits behind head on jazzy.

Files with missing lines	Patch %	Lines
...ments/include/dc_measurements/mission_registry.hpp	76.00%	6 Missing ⚠️
...asurements/include/dc_measurements/measurement.hpp	80.00%	3 Missing ⚠️
...s/include/dc_measurements/pending_record_queue.hpp	84.62%	2 Missing ⚠️
...ements/include/dc_measurements/mission_outcome.hpp	90.91%	1 Missing ⚠️
dc_measurements/plugins/measurements/fault.cpp	66.67%	1 Missing ⚠️
...ments/plugins/measurements/ros2_control_status.cpp	66.67%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##            jazzy     #418      +/-   ##
==========================================
- Coverage   70.80%   70.60%   -0.19%     
==========================================
  Files         114      119       +5     
  Lines        7440     7363      -77     
==========================================
- Hits         5267     5198      -69     
+ Misses       2173     2165       -8     

Flag	Coverage Δ
cpp-jazzy	70.60% <91.03%> (-0.19%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 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.