feat(climate): follow WHO 4 dimension 7 zone state on Home+Control plants (#429)

[GreenGrassBlueOcean]

Part of #429. Needs OpenWebNet-HA/OWNd#60 (fixes OWNd#58) to take effect; until the OWNd pin includes it, nothing changes.

Problem

On MyHomeServer1 + Home+Control plants, the zone's mode and setpoint only travel as WHO 4 dimension 7 (*#4*Z*7*CONTEXT*STATE[*TTTT]##). The reply to *#4*Z## never carries it (seen on both #429 plants). OWNd 2.0.0b8 parses dimension 7 with no message type, so the climate entity never saw it. The nominal setpoint also went stale, because the dim-12 *T*3 reply only updates the shown target (#454).

Change: climate.py

Handles MESSAGE_TYPE_ZONE_STATE (OWNd#60):

Dimension 7 state	Mode	Setpoint
setpoint	heating → HEAT, cooling → COOL, automatic → AUTO (if the zone supports it); generic keeps the mode	becomes the nominal target; shown target = setpoint + local offset
comfort / eco	same as setpoint	unchanged
protection / off	OFF	nominal setpoint kept for HEAT/COOL to restore, as WHAT 102/202 already do (#383)
unknown	unchanged	unchanged

The message type and value names are spelled out locally until the OWNd pin exports them.

Traces and golden samples from #429

The reporters' traces are stored verbatim in tests/fixtures/traces/issue_429, with a README recording who recorded what and what they did at each time. The privacy checker finds no personal data in them.

• @TheDarkWizard (MyHomeServer1 2.87.13, 4 KM4691 zones, zone 4 heating-only), 8 card exports:
  • OFF → manual → OFF
  • setpoint steps 29 / 23 / 33 °C
  • summer scenario, winter command, heat ↔ cool switches
  • Away (22.5 → 26.5 → 21 °C)
  • a restart
  • a manual override and a schedule step to Summer ECO
• @xtimmy86x (F454 on a MyHomeServer1 plant, 7 zones):
  • the 13 h Night → Comfort → Eco program trace
  • the follow-up showing that a setpoint written from Home Assistant takes the zone out of MyHomeServer1's program: the next #7 writes go to zones 2–7 only, and zone 1 keeps 20 °C

tests/test_issue_429_zone_state.py replays each trace through one climate entity per zone and checks the mode and nominal setpoint at the reporters' times. Findings pinned by the tests:

• a 35.0 °C protection value arrives just before the real setpoint when a zone leaves OFF;
• Away and schedule steps are plain setpoints, with no preset on the bus;
• status replies after a restart carry no dimension 7.

The 9 dimension-7 replays are a strict xfail while the installed OWNd drops dimension 7, following the same detect-don't-pin pattern as the other OWNd markers.

The golden corpus gets the same 9 verbatim community-plant-capture fixtures as OWNd#60, with provenance in SOURCE.yaml. corpus.json is regenerated (179 fixtures; the README count was stale at 58).

Unit tests in test_component_climate.py cover the cases the traces don't reach: comfort/eco without a temperature, generic context, a context the zone doesn't support, a setpoint while OFF, unknown values, and the local offset.

Tests

• Pinned OWNd 2.0.0b8: 2044 passed, 15 xfailed (9 new plus 6 that already existed).
• Against OWNd#60: 2059 passed.
• ruff and mypy --strict on climate.py are clean.

Checklist

• V2 Architecture: New entities implement handle_event() and do not manually subscribe via async_dispatcher_connect. (No new entities; the existing climate handle_event gains a branch.)

After OWNd#60 is released

Bump the OWNd pin. The 9 strict xfails in test_issue_429_zone_state.py then pass and fail the build, so drop needs_dimension_7 in the same change.

🤖 Generated with Claude Code

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!