Buckets:
AnonymousSpottedToad/candidate_bucket / codebase /python /als-apg_osprey /tests /simulation /test_apply_timezone.py
| """Regression (review Finding 1): seeded logbook timestamps land in the FACILITY | |
| zone, coherent with where telemetry ``at_time`` events are placed. | |
| ``apply_scenarios`` resolves each bundle entry's ``{days_ago, time}`` relative | |
| timestamp against the apply-time anchor; the anchor's tzinfo decides the zone the | |
| narrative time-of-day lands in. Before the fix the *default* anchor was | |
| ``datetime.now(UTC)``, so on a non-UTC facility a logbook entry documented at | |
| 03:20 landed at 03:20 UTC while the archiver spike it describes (a daily | |
| ``at_time`` event, placed facility-local by the simulation engine) landed at | |
| 03:20 facility-local — hours apart, breaking the correlation the scenario exists | |
| to exercise. | |
| DB-free: ``_seed_logbook`` is stubbed to capture the resolved entries, so the | |
| contract is pinned without a Postgres dependency and runs in the fast suite. | |
| """ | |
| from __future__ import annotations | |
| import shutil | |
| from pathlib import Path | |
| from zoneinfo import ZoneInfo | |
| import yaml | |
| from osprey.simulation.apply import apply_scenarios | |
| TEMPLATE_SIM = ( | |
| Path(__file__).resolve().parents[2] | |
| / "src/osprey/templates/apps/control_assistant/data/simulation" | |
| ) | |
| LA = ZoneInfo("America/Los_Angeles") # non-UTC facility; -7h (PDT) / -8h (PST) | |
| def _make_project(tmp_path: Path) -> Path: | |
| """Stage a minimal sim-backed project with a non-UTC facility timezone.""" | |
| sim_dst = tmp_path / "data" / "simulation" | |
| sim_dst.parent.mkdir(parents=True, exist_ok=True) | |
| shutil.copytree(TEMPLATE_SIM, sim_dst) | |
| config = { | |
| "control_system": { | |
| "connector": {"mock": {"simulation_file": "data/simulation/machine.json"}} | |
| }, | |
| # URI is never dialed: _seed_logbook is stubbed below. | |
| "ariel": {"database": {"uri": "postgresql://unused-mocked/none"}}, | |
| "system": {"timezone": "America/Los_Angeles"}, | |
| } | |
| (tmp_path / "config.yml").write_text(yaml.safe_dump(config)) | |
| return tmp_path | |
| def test_default_anchor_seeds_logbook_in_facility_zone(tmp_path, monkeypatch): | |
| project = _make_project(tmp_path) | |
| captured: dict = {} | |
| async def _fake_seed(ariel_config, entries): | |
| captured["entries"] = entries | |
| return len(entries), True | |
| monkeypatch.setattr("osprey.simulation.apply._seed_logbook", _fake_seed) | |
| # Pin the facility zone regardless of host config-loading quirks. | |
| monkeypatch.setattr("osprey.simulation.apply.get_facility_timezone", lambda: LA, raising=False) | |
| # now=None -> exercise the DEFAULT anchor (the path the bug lives on). | |
| apply_scenarios(project, ["rf-thermal"], now=None) | |
| entries = captured.get("entries") | |
| assert entries, "no logbook entries were seeded" | |
| # Every seeded timestamp is expressed in the facility zone (LA), not UTC. | |
| # A UTC anchor makes utcoffset() == 0 while LA is -7/-8h, so this is unequal | |
| # before the fix and equal after (DST-robust: compares the entry's own zone). | |
| for e in entries: | |
| ts = e["timestamp"] | |
| assert ts.tzinfo is not None, f"entry {e['entry_id']} has a naive timestamp" | |
| assert ts.utcoffset() == ts.astimezone(LA).utcoffset(), ( | |
| f"entry {e['entry_id']} not anchored in the facility zone: {ts.isoformat()}" | |
| ) | |
| # DEMO-026 is documented at days_ago=4, 03:20 — that time-of-day is the | |
| # FACILITY-LOCAL clock time, matching the archiver narrative it accompanies. | |
| demo026 = next(e["timestamp"] for e in entries if e["entry_id"] == "DEMO-026") | |
| local = demo026.astimezone(LA) | |
| assert (local.hour, local.minute) == (3, 20) | |
| # ...and NOT 03:20 UTC (the pre-fix placement). | |
| utc = demo026.astimezone(ZoneInfo("UTC")) | |
| assert (utc.hour, utc.minute) != (3, 20) | |
Xet Storage Details
- Size:
- 3.69 kB
- Xet hash:
- f219bcdf958e7026599fd8959700232f8329a2222a3e9fd407dfe28236e7137e
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.