Harden bus subscriber + readsb parser (PR #4/#5 review)

[kninetimmy]

Addresses the valid items from the Gemini reviews on the already-merged M1.2c (bus) and M2.1a (readsb parser) slices — now relevant because M2.1b put the readsb parser live on the bus.

Changes

bus/client.apply_payload — guard the handle() (state-apply) call. A bug or bad-state error while applying one record must not propagate and tear down the single subscriber ingest loop that feeds every source and client (PRD §37). It's now logged and dropped like a malformed payload. (This was the highest-value item in the review backlog.)

readsb parser robustness (PRD §17.2 "validate source responses"):

• _num rejects NaN/Infinity — Python's json accepts them by default, but they're invalid to re-emit and would crash a record's serialization on publish. Non-finite numeric fields normalize to None.
• _epoch_to_dt guards datetime.fromtimestamp against OverflowError/OSError/ValueError on out-of-range epochs, so a bad top-level now falls back to received_at instead of crashing the whole snapshot.
• absurd per-aircraft age no longer overflows timedelta; the track is kept, anchored at snapshot now.
• defensive isinstance(data, dict) guard at the parser entry.

Verification

Adds a test for each path. scripts/check.sh green (ruff + mypy strict + pytest, 67 passed).

Rejected from the review

Inferring on-ground from a "g" field — not a documented readsb aircraft.json field, so it stays out per the "no invented API fields" guardrail.

🤖 Generated with Claude Code

[gemini-code-assist]

Code Review

This pull request enhances the robustness and failure isolation of the readsb adapter and the bus client. Key changes include rejecting non-finite floats, handling potential exceptions (such as OverflowError, OSError, and ValueError) during timestamp and age calculations, and isolating exceptions raised during state application in the subscriber loop. The review feedback suggests further improving robustness by catching OverflowError during float conversion of extremely large integers in _num, and logging the specific exception details when dropping malformed payloads to aid in troubleshooting.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

Converting an extremely large integer (e.g., 10**309 or larger) to a float in Python raises an OverflowError rather than returning infinity. Since JSON allows arbitrarily large integers, if a client sends a very large integer for a numeric field (such as the top-level now field), _num will raise an unhandled OverflowError. Because the top-level now parsing is not wrapped in a try-except block in parse_aircraft_snapshot, this will crash the entire snapshot parsing, violating the failure isolation principle (PRD §37). Wrapping the float conversion in a try-except block to catch OverflowError prevents this crash.

Suggested change

	if isinstance(value, (int, float)):
	return float(value)
	result = float(value)
	return result if math.isfinite(result) else None
	if isinstance(value, (int, float)):
	try:
	result = float(value)
	except OverflowError:
	return None
	return result if math.isfinite(result) else None

[gemini-code-assist]

When dropping a malformed record payload due to validation or size errors, the actual exception details are discarded. Logging the exception message is highly beneficial for troubleshooting and debugging schema mismatches or corrupted payloads in production.

Suggested change

	except (ValidationError, RecordTooLargeError, ValueError):
	log.warning("dropping malformed record payload")
	return False
	except (ValidationError, RecordTooLargeError, ValueError) as exc:
	log.warning("dropping malformed record payload: %s", exc)
	return False