2026-07-04 22:22:33 -06:00
|
|
|
"""Phase-3 hydro (USGS NWIS) refactor tests.
|
|
|
|
|
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
Verifies the formatter+decider for the stream-gauge hazard (gating/hydro.py,
|
|
|
|
|
formatters/hydro.py), mirroring test_quake_refactor.py:
|
|
|
|
|
|
|
|
|
|
1. Golden: formatters.hydro.format() renders the expected wire, for a
|
|
|
|
|
stage-only crossing, a paired flow(00060)+stage(00065) reading, and every
|
|
|
|
|
threshold label.
|
|
|
|
|
|
|
|
|
|
2. Gate-sequence: an explicit `now`-timeline of readings driven through
|
|
|
|
|
gating.hydro.decide() (upward crossing broadcasts; same-rank + receding
|
|
|
|
|
suppress unless broadcast_on_recede).
|
|
|
|
|
|
|
|
|
|
The Central `nwis_handler` module (`_render()`, `handle_nwis()`) has been
|
|
|
|
|
deleted along with the rest of the Central NATS consumer path. Pure
|
|
|
|
|
old-vs-new parity assertions have been removed (original diffs are
|
|
|
|
|
preserved in git history); what remains asserts against hand-written
|
|
|
|
|
expected strings / broadcast outcomes.
|
|
|
|
|
|
|
|
|
|
decide() is READ-ONLY over gauge_readings by design (the append-only INSERT
|
|
|
|
|
was always caller-owned -- previously the Central handler, inline,
|
|
|
|
|
immediately after calling decide()). With the handler gone there is no
|
|
|
|
|
current producer for the "stream_flow" category in production (no native
|
|
|
|
|
env/ adapter emits it -- meshai.env.usgs.USGSStreamsAdapter is a separate,
|
|
|
|
|
older stream-gauge pipeline with different categories/schema). Tests below
|
|
|
|
|
that need prior-reading state seed gauge_readings directly via a local SQL
|
|
|
|
|
helper that mirrors the deleted handler's INSERT shape, so the gate logic
|
|
|
|
|
itself stays under direct, native-only test coverage.
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
The real registry / cutover key is "stream_flow" — the flat category the
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
Central nwis path used to produce for every central.hydro.* envelope.
|
2026-07-04 22:22:33 -06:00
|
|
|
"""
|
|
|
|
|
from __future__ import annotations
|
|
|
|
|
|
|
|
|
|
import pytest
|
|
|
|
|
|
|
|
|
|
from meshai.persistence import close_thread_connection, init_db
|
|
|
|
|
from meshai.persistence import db as persistence_db
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
from tests.harness.goldens import assert_byte_identical
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
_AT = 1_783_200_000.0 # pinned epoch (unused by hydro render/gate, kept for parity)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# ── DB fixture (same shape as test_nwis_handler.py) ──────────────────────────
|
|
|
|
|
|
|
|
|
|
@pytest.fixture
|
|
|
|
|
def mem_db(monkeypatch, tmp_path):
|
|
|
|
|
db_path = str(tmp_path / "hydro-refactor-test.sqlite")
|
|
|
|
|
monkeypatch.setenv("MESHAI_DB_PATH", db_path)
|
|
|
|
|
persistence_db._initialised.clear()
|
|
|
|
|
close_thread_connection()
|
|
|
|
|
conn = init_db()
|
|
|
|
|
yield conn
|
|
|
|
|
close_thread_connection()
|
|
|
|
|
persistence_db._initialised.discard(db_path)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def _make_fake_event(data: dict):
|
|
|
|
|
class _FakeEvent:
|
|
|
|
|
pass
|
|
|
|
|
e = _FakeEvent()
|
|
|
|
|
e.data = data
|
|
|
|
|
return e
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# ─────────────────────────────────────────────────────────────────────────────
|
|
|
|
|
# 1. Golden byte-identical — formatter reproduces _render() exactly
|
|
|
|
|
# ─────────────────────────────────────────────────────────────────────────────
|
|
|
|
|
|
|
|
|
|
class TestFormatterGolden:
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
"""formatters.hydro.format() renders the expected wire for canonical data."""
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
def _fmt_new(self, canonical: dict) -> str:
|
|
|
|
|
from meshai.notifications.formatters.hydro import format as hfmt
|
|
|
|
|
return hfmt(_make_fake_event(canonical), now=_AT, budget=140)
|
|
|
|
|
|
|
|
|
|
def test_stage_only_crossing_action(self):
|
|
|
|
|
"""Stage-only (00065) reading at action stage, no companion flow."""
|
|
|
|
|
canonical = {
|
|
|
|
|
"gauge_name": "Snake River at Heise",
|
|
|
|
|
"threshold_state": "action",
|
|
|
|
|
"stage_ft": 12.5,
|
|
|
|
|
"flow_cfs": None,
|
|
|
|
|
"unit": "ft",
|
|
|
|
|
"lat": 43.612,
|
|
|
|
|
"lon": -111.654,
|
|
|
|
|
}
|
|
|
|
|
new = self._fmt_new(canonical)
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
assert_byte_identical(
|
|
|
|
|
new, "🌊 New: Snake River at Heise: action stage 12.5 ft, @ 43.612,-111.654"
|
|
|
|
|
)
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
def test_paired_flow_and_stage(self):
|
|
|
|
|
"""00060 discharge back-looked onto a 00065 stage: flow segment present."""
|
|
|
|
|
canonical = {
|
|
|
|
|
"gauge_name": "Boise River",
|
|
|
|
|
"threshold_state": "flood_minor",
|
|
|
|
|
"stage_ft": 14.5,
|
|
|
|
|
"flow_cfs": 8400,
|
|
|
|
|
"unit": "ft",
|
|
|
|
|
"lat": 43.600,
|
|
|
|
|
"lon": -116.200,
|
|
|
|
|
}
|
|
|
|
|
new = self._fmt_new(canonical)
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
assert_byte_identical(
|
|
|
|
|
new,
|
|
|
|
|
"🌊 New: Boise River: minor flooding 14.5 ft, flow 8,400 cfs, @ 43.600,-116.200",
|
|
|
|
|
)
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
|
|
|
"state,label",
|
|
|
|
|
[
|
|
|
|
|
("action", "action stage"),
|
|
|
|
|
("flood_minor", "minor flooding"),
|
|
|
|
|
("flood_moderate", "moderate flooding"),
|
|
|
|
|
("flood_major", "major flooding"),
|
|
|
|
|
],
|
|
|
|
|
)
|
|
|
|
|
def test_every_threshold_label(self, state, label):
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
"""Each threshold_state maps to the correct label."""
|
2026-07-04 22:22:33 -06:00
|
|
|
canonical = {
|
|
|
|
|
"gauge_name": "Test Gauge",
|
|
|
|
|
"threshold_state": state,
|
|
|
|
|
"stage_ft": 20.0,
|
|
|
|
|
"flow_cfs": None,
|
|
|
|
|
"unit": "ft",
|
|
|
|
|
"lat": 44.0,
|
|
|
|
|
"lon": -114.0,
|
|
|
|
|
}
|
|
|
|
|
new = self._fmt_new(canonical)
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
assert_byte_identical(
|
|
|
|
|
new, f"🌊 New: Test Gauge: {label} 20.0 ft, @ 44.000,-114.000"
|
|
|
|
|
)
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
def test_missing_coords_drops_at_tail(self):
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
"""No coords → no @ segment."""
|
2026-07-04 22:22:33 -06:00
|
|
|
canonical = {
|
|
|
|
|
"gauge_name": "No Coords Gauge",
|
|
|
|
|
"threshold_state": "action",
|
|
|
|
|
"stage_ft": 10.0,
|
|
|
|
|
"flow_cfs": None,
|
|
|
|
|
"unit": "ft",
|
|
|
|
|
"lat": None,
|
|
|
|
|
"lon": None,
|
|
|
|
|
}
|
|
|
|
|
new = self._fmt_new(canonical)
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
assert_byte_identical(new, "🌊 New: No Coords Gauge: action stage 10.0 ft")
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
|
|
|
|
|
# ─────────────────────────────────────────────────────────────────────────────
|
|
|
|
|
# 2. Gate-sequence parity — old handle_nwis vs new gating.hydro.decide()
|
|
|
|
|
# ─────────────────────────────────────────────────────────────────────────────
|
|
|
|
|
|
|
|
|
|
def _nwis_env(*, site_id="USGS-13186000", parameter_code="00065", value=13.0,
|
|
|
|
|
unit="ft", time_iso="2026-06-05T15:00:00Z",
|
|
|
|
|
lat=43.612, lon=-111.654, envelope_id=None):
|
|
|
|
|
envelope_id = envelope_id or f"nwis_{site_id}_{time_iso}"
|
|
|
|
|
return {
|
|
|
|
|
"id": envelope_id,
|
|
|
|
|
"subject": f"central.hydro.{parameter_code}.usgs.{site_id}.us.id",
|
|
|
|
|
"data": {
|
|
|
|
|
"id": envelope_id, "adapter": "nwis",
|
|
|
|
|
"category": f"hydro.{parameter_code}", "severity": 0,
|
|
|
|
|
"geo": {"centroid": [lon, lat], "primary_region": "US-ID"},
|
|
|
|
|
"data": {
|
|
|
|
|
"id": envelope_id,
|
|
|
|
|
"monitoring_location_id": site_id,
|
|
|
|
|
"parameter_code": parameter_code,
|
|
|
|
|
"time": time_iso,
|
|
|
|
|
"value": value,
|
|
|
|
|
"unit_of_measure": unit,
|
|
|
|
|
"latitude": lat, "longitude": lon,
|
|
|
|
|
},
|
|
|
|
|
},
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def _parse_iso_epoch(s):
|
|
|
|
|
from datetime import datetime
|
|
|
|
|
return int(datetime.fromisoformat(s.replace("Z", "+00:00")).timestamp())
|
|
|
|
|
|
|
|
|
|
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
def _insert_reading(conn, *, site_id, gauge_name, value, unit,
|
|
|
|
|
threshold_state, flow_cfs, reading_time, lat, lon):
|
|
|
|
|
"""Directly seed a gauge_readings row.
|
|
|
|
|
|
|
|
|
|
Mirrors the schema the deleted Central nwis_handler used to INSERT
|
|
|
|
|
inline, immediately after calling decide(). decide() is read-only over
|
|
|
|
|
this table by design (see gating/hydro.py docstring) -- the INSERT was
|
|
|
|
|
always caller-owned, so tests seed state directly instead of reaching
|
|
|
|
|
into the deleted handler.
|
|
|
|
|
"""
|
|
|
|
|
conn.execute(
|
|
|
|
|
"INSERT INTO gauge_readings(site_id, gauge_name, reading_value, "
|
|
|
|
|
"reading_unit, threshold_state, flow_cfs, reading_time, lat, lon) "
|
|
|
|
|
"VALUES (?,?,?,?,?,?,?,?,?)",
|
|
|
|
|
(site_id, gauge_name, value, unit, threshold_state, flow_cfs,
|
|
|
|
|
reading_time, lat, lon),
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
2026-07-04 22:22:33 -06:00
|
|
|
class TestGateSequenceParity:
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
"""gating.hydro.decide() broadcast/suppress across a reading timeline."""
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
@pytest.fixture(autouse=True)
|
|
|
|
|
def _db(self, mem_db):
|
|
|
|
|
self.db = mem_db
|
|
|
|
|
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
def _canonical(self, fixture):
|
|
|
|
|
"""Build canonical data from a Central-style fixture (as the deleted
|
|
|
|
|
handler used to) via the still-live idaho_gauge_sites helpers."""
|
2026-07-04 22:22:33 -06:00
|
|
|
from meshai.central.idaho_gauge_sites import (
|
|
|
|
|
compute_threshold_state, lookup_site, normalize_site_id,
|
|
|
|
|
)
|
|
|
|
|
env = fixture["envelope"]
|
|
|
|
|
d = env["data"]["data"]
|
|
|
|
|
raw_site = d.get("monitoring_location_id")
|
|
|
|
|
site_id = normalize_site_id(raw_site)
|
|
|
|
|
site_meta = lookup_site(raw_site)
|
|
|
|
|
pc = d.get("parameter_code")
|
|
|
|
|
value = float(d.get("value"))
|
|
|
|
|
reading_time = _parse_iso_epoch(d.get("time"))
|
|
|
|
|
stage_ft = value if pc == "00065" else None
|
|
|
|
|
flow_cfs = value if pc == "00060" else None
|
|
|
|
|
threshold_state = "normal"
|
|
|
|
|
if pc == "00065":
|
|
|
|
|
threshold_state = compute_threshold_state(stage_ft, site_meta)
|
|
|
|
|
canonical = {
|
|
|
|
|
"site_id": site_id,
|
|
|
|
|
"gauge_name": site_meta["gauge_name"],
|
|
|
|
|
"stage_ft": stage_ft,
|
|
|
|
|
"flow_cfs": flow_cfs,
|
|
|
|
|
"unit": d.get("unit_of_measure"),
|
|
|
|
|
"threshold_state": threshold_state,
|
|
|
|
|
"reading_time": reading_time,
|
|
|
|
|
"lat": d.get("latitude"),
|
|
|
|
|
"lon": d.get("longitude"),
|
|
|
|
|
"parameter_code": pc,
|
|
|
|
|
}
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
return canonical, value
|
|
|
|
|
|
|
|
|
|
def _decide(self, fixture, *, now):
|
|
|
|
|
"""decide() only (no persist) — for assertions on a single reading."""
|
|
|
|
|
from meshai.notifications.gating.hydro import decide
|
|
|
|
|
canonical, _value = self._canonical(fixture)
|
2026-07-04 22:22:33 -06:00
|
|
|
return decide(canonical, source="nwis", now=float(now))
|
|
|
|
|
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
def _decide_and_persist(self, fixture, *, now):
|
|
|
|
|
"""decide() then INSERT the resolved reading — mirrors the deleted
|
|
|
|
|
handler's decide-then-insert ordering, so later steps in a sequence
|
|
|
|
|
see accumulated prior state exactly as production would."""
|
|
|
|
|
from meshai.notifications.gating.hydro import decide
|
|
|
|
|
canonical, value = self._canonical(fixture)
|
|
|
|
|
gate = decide(canonical, source="nwis", now=float(now))
|
|
|
|
|
threshold_state = gate.data_patch.get("threshold_state", canonical["threshold_state"])
|
|
|
|
|
stage_ft = gate.data_patch.get("stage_ft", canonical["stage_ft"])
|
|
|
|
|
_insert_reading(
|
|
|
|
|
self.db,
|
|
|
|
|
site_id=canonical["site_id"], gauge_name=canonical["gauge_name"],
|
|
|
|
|
value=value, unit=canonical["unit"], threshold_state=threshold_state,
|
|
|
|
|
flow_cfs=canonical["flow_cfs"], reading_time=canonical["reading_time"],
|
|
|
|
|
lat=canonical["lat"], lon=canonical["lon"],
|
|
|
|
|
)
|
|
|
|
|
return gate
|
|
|
|
|
|
2026-07-04 22:22:33 -06:00
|
|
|
def test_gate_sequence_matches(self):
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
"""Timeline of Heise readings: decide() agrees with expectations at every step.
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
Heise (USGS-13186000): action=12.0ft.
|
|
|
|
|
[0] 8.0 ft normal (first reading, no prior) → suppress
|
|
|
|
|
[1] 12.5 ft action (normal → action upward) → broadcast
|
|
|
|
|
[2] 12.8 ft action (action → action same rank) → suppress
|
|
|
|
|
[3] 14.5 ft f_minor (action → flood_minor upward) → broadcast
|
|
|
|
|
[4] 8.0 ft normal (flood_minor → normal receding) → suppress
|
|
|
|
|
"""
|
|
|
|
|
base = _parse_iso_epoch("2026-06-05T10:00:00Z")
|
|
|
|
|
specs = [
|
|
|
|
|
(8.0, "2026-06-05T10:00:00Z", "s0"),
|
|
|
|
|
(12.5, "2026-06-05T10:15:00Z", "s1"),
|
|
|
|
|
(12.8, "2026-06-05T10:30:00Z", "s2"),
|
|
|
|
|
(14.5, "2026-06-05T10:45:00Z", "s3"),
|
|
|
|
|
(8.0, "2026-06-05T11:00:00Z", "s4"),
|
|
|
|
|
]
|
|
|
|
|
ordered = [
|
|
|
|
|
{"envelope": _nwis_env(value=v, time_iso=t, envelope_id=eid)}
|
|
|
|
|
for (v, t, eid) in specs
|
|
|
|
|
]
|
|
|
|
|
timeline = [float(base + i * 900) for i in range(len(specs))]
|
|
|
|
|
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
results = [
|
|
|
|
|
self._decide_and_persist(fx, now=t)
|
|
|
|
|
for fx, t in zip(ordered, timeline)
|
|
|
|
|
]
|
|
|
|
|
|
|
|
|
|
assert results[0].broadcast is False, "normal first reading suppressed"
|
|
|
|
|
assert results[1].broadcast is True, "normal→action broadcasts"
|
|
|
|
|
assert results[2].broadcast is False, "action→action suppressed"
|
|
|
|
|
assert results[3].broadcast is True, "action→flood_minor broadcasts"
|
|
|
|
|
assert results[4].broadcast is False, "receding suppressed (no toggle)"
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
def test_00060_backlook_inherits_stage_band(self, mem_db):
|
|
|
|
|
"""A 00060 discharge reading inherits the last 00065 stage band.
|
|
|
|
|
|
|
|
|
|
Seed an action-stage 00065 reading, then feed a 00060 discharge: the
|
|
|
|
|
decider's back-look must resolve threshold_state=action + the prior
|
|
|
|
|
stage_ft, and (same rank as the seeded action) suppress the discharge.
|
|
|
|
|
"""
|
|
|
|
|
env_stage = _nwis_env(parameter_code="00065", value=12.5,
|
|
|
|
|
time_iso="2026-06-05T10:00:00Z", envelope_id="seed")
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
self._decide_and_persist({"envelope": env_stage}, now=1_000_000)
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
# Now decide on a 00060 discharge — should back-look the action band.
|
|
|
|
|
env_flow = _nwis_env(parameter_code="00060", value=8400, unit="ft^3/s",
|
|
|
|
|
time_iso="2026-06-05T10:05:00Z", envelope_id="q")
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
gate = self._decide({"envelope": env_flow}, now=1_000_300)
|
2026-07-04 22:22:33 -06:00
|
|
|
assert gate.data_patch["threshold_state"] == "action"
|
|
|
|
|
assert gate.data_patch["stage_ft"] == 12.5
|
|
|
|
|
# action → action (same rank) → suppress
|
|
|
|
|
assert gate.broadcast is False
|
|
|
|
|
|
|
|
|
|
def test_recede_toggle_enables_broadcast(self, mem_db):
|
|
|
|
|
"""With broadcast_on_recede set, a receding crossing broadcasts."""
|
|
|
|
|
from meshai.adapter_config._accessor import set_runtime_override, _overrides
|
|
|
|
|
# Seed an action reading.
|
|
|
|
|
env_high = _nwis_env(parameter_code="00065", value=12.5,
|
|
|
|
|
time_iso="2026-06-05T10:00:00Z", envelope_id="hi")
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
self._decide_and_persist({"envelope": env_high}, now=1_000_000)
|
2026-07-04 22:22:33 -06:00
|
|
|
|
|
|
|
|
# Force the recede toggle on for the decision only (runtime override,
|
|
|
|
|
# since adapter_config accessors are read-only).
|
|
|
|
|
set_runtime_override("usgs_nwis", "broadcast_on_recede", True)
|
|
|
|
|
try:
|
|
|
|
|
env_low = _nwis_env(parameter_code="00065", value=8.0,
|
|
|
|
|
time_iso="2026-06-05T11:00:00Z", envelope_id="lo")
|
chore: excise the dead Central NATS consumer path (-11,328 LOC) (#144)
* chore: excise the dead Central NATS consumer path
Central was retired and its database dropped 2026-07-15; its NATS broker no
longer exists. Verified against the live CT108 deployment: all 12 adapters
run feed_source=native, zero on central, and central.enabled is False
(default, never overridden). The consumer and its handlers were unreachable.
Removed:
- central/consumer.py and 6 dead handlers (nws, quake, swpc, nwis, avy,
incident) -- their handle_* entrypoints were reachable only from the
consumer's dispatch
- the Central wiring in main.py (init, guarded start, retry loop, stop path)
- the dead config surface: CentralConsumerConfig, EnvironmentalConfig.central,
adapter_config ("central","severity_thresholds") and its display block
- the nats-py dependency (consumer.py was its only importer)
- 19 test files that exercised only the dead path
KEPT -- these live under central/ but are imported directly by native
adapters, and deleting them would break production:
- wfigs_handler.py: firms_handler._handle_pass_boundary() calls its _render()
on the live FIRMS growth-fire path (env/firms.py -> ingest_hotspot_pixel)
- firms_handler, satpass_handler, tle_handler: split files whose handle_*
entrypoints are dead but whose engines are live. Left intact; splitting
them is separate work.
- pass_predictor, budget, idaho_gauge_sites: fully live.
The usgs_quake keys global_mag_floor / regional_mag_floor / regional_centroid
/ regional_radius_mi / broadcast_pager_alerts are NOT removed despite comments
labelling them "CENTRAL-PATH ONLY" -- notifications/gating/quake.py reads them
unconditionally in the native path. Those comments are corrected separately.
Test-count note: the suite drops ~425 tests. Most were migration PARITY tests
whose sole purpose was proving the native rewrite byte-matched the Central
handler (golden byte-parity, cross-source identity, gate-sequence replay).
With the handler deleted there is nothing left to compare against, so they
cannot exist. Native-only tests were kept and reworked where a test reached
for a central symbol incidentally. This is a real coverage loss, accepted
deliberately: the parity harness proved the refactor faithful, and git
history preserves the originals.
Suite: 1984 passed, 6 failed -- the same 6 pre-existing failures as main
(stale SCHEMA_VERSION x3, expired TLE fixtures x2, one order-dependent),
all being fixed on fix/green-test-suite. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(nws): restore native-only golden coverage for the wire formatter
Commit ca751fb5 deleted the Central nws_handler parity harness along with
the handler itself, which took the ONLY tests that pinned formatters.nws
.format()'s literal wire output. Gate-sequence and schema-conformance
tests already survived natively; the formatter's actual rendered text did
not have any native-only regression net.
Add TestFormatterGolden to test_nws_refactor.py: 3 real-fixture cases plus
6 hand-built pathological cases mined from the deleted test_nws_handler.py
(SVR path-sampling, the "no dangling separator" regression, TOR on-ground
vs radar-indicated, FFW flood-cause detection). Every literal was verified
by temporarily restoring the pre-excision central.nws_handler._render()
from git history (ca751fb5^) in a throwaway, uncommitted script, confirming
byte-identical output against the current native format() for all 37 real
fixtures (nws/ + nws_last/) and all 9 pathological cases, then pinning the
confirmed-matching string as the literal -- not a blind snapshot of
current behavior.
quake/swpc/avalanche/hydro/incident/fire were checked and already carry
equivalent native-only golden coverage (added directly in ca751fb5), so no
changes were needed there.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Matt Johnson <mj@k7zvx.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-17 14:03:40 -06:00
|
|
|
gate = self._decide({"envelope": env_low}, now=1_003_600)
|
2026-07-04 22:22:33 -06:00
|
|
|
finally:
|
|
|
|
|
_overrides.pop(("usgs_nwis", "broadcast_on_recede"), None)
|
|
|
|
|
assert gate.broadcast is True, "receding must broadcast when toggle is on"
|
|
|
|
|
assert gate.data_patch["threshold_state"] == "normal"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# ─────────────────────────────────────────────────────────────────────────────
|
|
|
|
|
# 3. Registration — stream_flow resolves to hydro formatter + decider
|
|
|
|
|
# ─────────────────────────────────────────────────────────────────────────────
|
|
|
|
|
|
|
|
|
|
class TestRegistration:
|
|
|
|
|
def test_stream_flow_formatter_registered(self):
|
|
|
|
|
from meshai.notifications.formatters import get_formatter
|
|
|
|
|
from meshai.notifications.formatters.hydro import format as hfmt
|
|
|
|
|
assert get_formatter("stream_flow") is hfmt
|
|
|
|
|
|
|
|
|
|
def test_stream_flow_decider_registered(self):
|
|
|
|
|
from meshai.notifications.gating import get_decider
|
|
|
|
|
from meshai.notifications.gating.hydro import decide
|
|
|
|
|
assert get_decider("stream_flow") is decide
|