forecastsolar: wrap HA-ML WebSocket failures as ProviderError

[MaStr]

ForecastSolarHomeAssistantML.get_raw_data_from_provider() let DNS,
connection, and auth failures from the WebSocket client escape
unwrapped. ForecastSolarBaseclass.refresh_data() only catches
(ConnectionError, TimeoutError, ProviderError), so these exceptions
bypassed the cache fallback entirely (same class of bug as #408 in
FCSolar, but for the websockets-based HA-ML provider).

Wrap OSError, WebSocketException, and RuntimeError from the fetch as
ProviderError so cached forecast data remains usable until its TTL
expires.

[Copilot]

Pull request overview

This PR improves reliability of the Home Assistant ML solar forecast provider by ensuring WebSocket DNS/connection/auth failures are surfaced as ProviderError, allowing ForecastSolarBaseclass.refresh_data() to fall back to cached raw data until TTL expiry.

Changes:

• Wrap OSError, WebSocketException, and RuntimeError from the HA-ML WebSocket fetch path as ProviderError.
• Add regression tests covering DNS failure, WebSocket failure, auth failure, and cache fallback behavior on connection failure.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
tests/test_forecast_solar_homeassistant_ml.py	Adds regression tests asserting network/auth failures are wrapped and cache fallback remains effective.
src/batcontrol/forecastsolar/forecast_homeassistant_ml.py	Wraps WebSocket-related failures as ProviderError so baseclass error handling can use cached data.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Suppressed comments (1)

src/batcontrol/forecastsolar/forecast_homeassistant_ml.py:344

• Catching RuntimeError here also wraps the "event loop is already running" error from run_until_complete(). That turns a programming/usage error into a ProviderError, which can silently trigger cache fallback and make the underlying issue hard to diagnose. Guard against a running loop before calling run_until_complete (or avoid wrapping that specific RuntimeError).

        try:
            loop = asyncio.get_event_loop()
        except RuntimeError:
            loop = asyncio.new_event_loop()
            asyncio.set_event_loop(loop)

        try:
            return loop.run_until_complete(self._fetch_entity_state_async())
        except (OSError, WebSocketException, RuntimeError) as e:
            logger.error(
                'HomeAssistant WebSocket request failed for entity %s: %s',
                self.entity_id, e, exc_info=True)
            raise ProviderError(
                f'HomeAssistant WebSocket request failed for entity {self.entity_id}: {e}'
            ) from e