feat(worker): chunk the discovery sweep (one chunk per loop iteration)

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Jonathan
2026-07-13 00:57:00 +02:00
parent 5351cd029d
commit a67c552f29
2 changed files with 72 additions and 30 deletions
+42 -16
View File
@@ -1,7 +1,7 @@
from lyra_worker import main
from lyra_worker.adapters.fakes import FakeMbBrowser, FakeSimilaritySource
from lyra_worker.discovery import DiscoveryConfig
from lyra_worker.main import _run_discovery, maybe_run_discovery
from lyra_worker.discovery import DiscoveryConfig, DiscoveryResult
from lyra_worker.main import maybe_run_discovery
from lyra_worker.similarity.base import SimilarArtist
from tests.conftest import insert_watched_artist
@@ -23,25 +23,49 @@ def _get(conn, key):
return row[0] if row else None
def test_run_discovery_writes_result_and_clears_flag(conn):
insert_watched_artist(conn, mbid="s1", name="Seed")
def test_maybe_run_discovery_drains_across_chunks_and_finishes(conn):
# The sweep must advance one chunk (discover.chunkSize seeds) per call, clear
# discover.requested exactly once at start, and finally drain on the 0-seed chunk:
# writing discover.result and clearing discover.inProgress.
for i in range(3):
insert_watched_artist(conn, mbid=f"s{i}", name=f"Seed{i}")
_set(conn, "discover.requested", "true")
src = FakeSimilaritySource(similar={"s1": [SimilarArtist("c1", "Cand", 0.9)]})
_run_discovery(conn, [src], FakeMbBrowser())
assert _get(conn, "discover.requested") == "false" # flag cleared
assert "artists" in (_get(conn, "discover.result") or "") # summary written
with conn.cursor() as cur:
cur.execute('SELECT count(*) FROM "DiscoverySuggestion" WHERE kind = \'artist\'')
assert cur.fetchone()[0] == 1
_set(conn, "discover.chunkSize", "2")
src = FakeSimilaritySource(
similar={f"s{i}": [SimilarArtist(f"c{i}", f"C{i}", 0.9)] for i in range(3)}
)
cfg = {"discover.enabled": "false", "discover.requested": "true", "discover.chunkSize": "2"}
tick = 0.0
# iteration 1: starts sweep, processes 2 seeds, still in progress
tick = maybe_run_discovery(
conn, [src], FakeMbBrowser(), cfg, DiscoveryConfig.from_config(cfg), 1.0, tick
)
assert _get(conn, "discover.inProgress") == "true"
assert _get(conn, "discover.requested") == "false" # cleared at sweep start
# refresh the snapshot the way run_forever's loop does (re-reads config each iteration)
cfg2 = {**cfg, "discover.requested": "false", "discover.inProgress": "true",
"discover.progress": _get(conn, "discover.progress")}
# iteration 2: processes the last seed, still in progress
maybe_run_discovery(
conn, [src], FakeMbBrowser(), cfg2, DiscoveryConfig.from_config(cfg2), 2.0, tick
)
assert _get(conn, "discover.inProgress") == "true"
cfg3 = {**cfg2, "discover.progress": _get(conn, "discover.progress")}
# iteration 3: 0 seeds left -> drains, writes result, clears inProgress
maybe_run_discovery(
conn, [src], FakeMbBrowser(), cfg3, DiscoveryConfig.from_config(cfg3), 3.0, tick
)
assert _get(conn, "discover.inProgress") == "false"
assert _get(conn, "discover.progress") == ""
assert "artists" in (_get(conn, "discover.result") or "")
def test_maybe_run_discovery_runs_once_when_due_and_requested(conn, monkeypatch):
# A scheduled tick that is due AND a pending one-shot request in the same loop
# iteration must trigger discovery exactly once, not twice.
calls = []
monkeypatch.setattr(main, "_run_discovery", lambda *a, **k: calls.append(1))
monkeypatch.setattr(main, "_run_discovery",
lambda *a, **k: (calls.append(1), DiscoveryResult(0, 0, seeds=0))[1])
config = {"discover.enabled": "true", "discover.requested": "true"}
dcfg = DiscoveryConfig.from_config(config)
new_tick = maybe_run_discovery(
@@ -55,7 +79,8 @@ def test_maybe_run_discovery_requested_when_disabled_does_not_advance_tick(conn,
# A one-shot request runs even when discovery is disabled, but must not advance the
# schedule tick (there is no active schedule).
calls = []
monkeypatch.setattr(main, "_run_discovery", lambda *a, **k: calls.append(1))
monkeypatch.setattr(main, "_run_discovery",
lambda *a, **k: (calls.append(1), DiscoveryResult(0, 0, seeds=0))[1])
config = {"discover.enabled": "false", "discover.requested": "true"}
dcfg = DiscoveryConfig.from_config(config)
new_tick = maybe_run_discovery(
@@ -67,7 +92,8 @@ def test_maybe_run_discovery_requested_when_disabled_does_not_advance_tick(conn,
def test_maybe_run_discovery_idle_does_nothing(conn, monkeypatch):
calls = []
monkeypatch.setattr(main, "_run_discovery", lambda *a, **k: calls.append(1))
monkeypatch.setattr(main, "_run_discovery",
lambda *a, **k: (calls.append(1), DiscoveryResult(0, 0, seeds=0))[1])
config = {"discover.enabled": "true", "discover.requested": "false"}
dcfg = DiscoveryConfig.from_config(config)
# enabled but not yet due (tick was recent relative to now)