fix: avoid repeated task and source scans on dashboard
Some checks failed
Mostovik Backend CI/CD / Tests and lint (pull_request) Successful in 3m16s
Mostovik Backend CI/CD / Build linux/amd64 release images (pull_request) Has been skipped
Mostovik Backend CI/CD / Deploy dev (pull_request) Has been skipped
Mostovik Backend CI/CD / Deploy and verify internal main (pull_request) Has been skipped
Mostovik Backend CI/CD / Tests and lint (push) Failing after 3s
Mostovik Backend CI/CD / Build linux/amd64 release images (push) Has been skipped
Mostovik Backend CI/CD / Deploy dev (push) Has been skipped
Mostovik Backend CI/CD / Deploy and verify internal main (push) Has been skipped

This commit was merged in pull request #47.
This commit is contained in:
Aleksandr Meshchryakov
2026-10-01 13:16:11 +02:00
parent 887442e8e3
commit fedd2e00a3
7 changed files with 239 additions and 37 deletions

View File

@@ -2,6 +2,7 @@
from datetime import timedelta
from unittest.mock import MagicMock
from uuid import UUID
import pytest
from apps.core.models import BackgroundJob, JobStatus
@@ -145,3 +146,53 @@ def test_latest_task_does_not_reactivate_a_stale_job(status, time_field):
**{time_field: timezone.now() - timedelta(minutes=15)}
)
assert SourceCardService.get_card(definition.slug)["active_tasks"] == []
@pytest.mark.django_db
def test_latest_job_uses_created_time_then_uuid_and_keeps_older_active_jobs():
definition = SourceCardService.get_definition("arbitration-cases")
jobs = [
BackgroundJob.objects.create(
id=UUID(int=index),
task_id=f"ordered-{index}",
task_name=definition.task_names[0],
status=JobStatus.SUCCESS,
)
for index in (1, 2, 3, 4)
]
now = timezone.now()
BackgroundJob.objects.filter(pk__in=[job.pk for job in jobs]).update(
created_at=now - timedelta(minutes=2)
)
# UUID 4 is older; UUID 2 wins the timestamp tie with UUID 1.
BackgroundJob.objects.filter(pk__in=[jobs[0].pk, jobs[1].pk]).update(created_at=now)
BackgroundJob.objects.filter(pk=jobs[2].pk).update(status=JobStatus.STARTED)
active, refresh = SourceCardService._build_task_maps((definition,))
assert [job["task_id"] for job in active[definition.slug]] == ["ordered-3"]
assert [job["task_id"] for job in refresh[definition.slug]] == [
"ordered-2",
"ordered-3",
]
@pytest.mark.django_db
def test_latest_job_is_chosen_before_filtering_source_card_metadata():
definition = SourceCardService.get_definition("arbitration-cases")
older = BackgroundJob.objects.create(
task_id="old-card-job",
task_name="custom.shared.task",
status=JobStatus.SUCCESS,
meta={"source_card": definition.slug},
)
newer = BackgroundJob.objects.create(
task_id="new-other-card-job",
task_name=older.task_name,
status=JobStatus.SUCCESS,
meta={"source_card": "bankruptcy-procedures"},
)
BackgroundJob.objects.filter(pk=older.pk).update(
created_at=newer.created_at - timedelta(minutes=1)
)
active, refresh = SourceCardService._build_task_maps((definition,))
assert active == {}
assert refresh == {}

View File

@@ -0,0 +1,61 @@
"""Snapshot counts avoid rescanning staged data when a saved total exists."""
from datetime import timedelta
import pytest
from apps.parsers.models import ParserLoadLog, ParserSourceArtifact, ParserStagedRecord
from apps.parsers.snapshot_metadata import attach_snapshot_artifacts
from django.db import connection
from django.test.utils import CaptureQueriesContext
from organizations.models import Organization
@pytest.mark.django_db
@pytest.mark.parametrize("saved_total", [0, 7, None, -1, "7"])
def test_saved_total_or_distinct_published_organizations(saved_total):
log = ParserLoadLog.objects.create(source="sro_membership_check", batch_id=1)
artifact = ParserSourceArtifact.objects.create(
source=log.source,
load_batch=log.batch_id,
metadata={"snapshot_organizations_count": saved_total},
)
organization = Organization.objects.create(name="Published fixture")
for row, disposition in enumerate(["published", "published", "quarantined"]):
ParserStagedRecord.objects.create(
artifact=artifact,
row_number=row + 1,
organization=organization,
disposition=disposition,
)
ParserStagedRecord.objects.create(
artifact=artifact, row_number=4, disposition="published", organization=None
)
with CaptureQueriesContext(connection) as queries:
attach_snapshot_artifacts([log, log])
cached = log._source_artifact_cache
assert cached.pk == artifact.pk
has_saved_total = isinstance(saved_total, int) and saved_total >= 0
assert cached.published_organizations_count == (
saved_total if has_saved_total else 1
)
staged_queries = [q for q in queries if "parsers_staged_record" in q["sql"]]
assert len(staged_queries) == (0 if has_saved_total else 1)
@pytest.mark.django_db
def test_latest_artifact_and_missing_load_keep_their_own_counts():
log = ParserLoadLog.objects.create(source="sro_membership_check", batch_id=1)
missing = ParserLoadLog.objects.create(source=log.source, batch_id=2)
old = ParserSourceArtifact.objects.create(source=log.source, load_batch=1)
latest = ParserSourceArtifact.objects.create(
source=log.source, load_batch=1, metadata={"snapshot_organizations_count": 9}
)
ParserSourceArtifact.objects.filter(pk=old.pk).update(
created_at=latest.created_at - timedelta(seconds=1)
)
with CaptureQueriesContext(connection) as queries:
attach_snapshot_artifacts([log, missing])
assert log._source_artifact_cache.pk == latest.pk
assert log._source_artifact_cache.published_organizations_count == 9
assert missing._source_artifact_cache is None
assert not any("parsers_staged_record" in q["sql"] for q in queries)

View File

@@ -20,6 +20,8 @@ from apps.parsers.models import (
ProcurementRecord,
)
from django.core.files.uploadedfile import SimpleUploadedFile
from django.db import connection
from django.test.utils import CaptureQueriesContext
from django.urls import reverse
from openpyxl import Workbook
from organizations.models import Organization, OrganizationSourceRecord
@@ -697,7 +699,17 @@ class ParsersViewSetTest(APITestCase):
)
self.client.force_authenticate(self.user)
response = self.client.get("/api/v1/parsers/dashboard/")
with CaptureQueriesContext(connection) as queries:
response = self.client.get("/api/v1/parsers/dashboard/")
source_scans = [
query["sql"]
for query in queries
if "SELECT DISTINCT" in query["sql"]
and "organizations_source_record" in query["sql"]
]
self.assertTrue(source_scans)
self.assertEqual(len(source_scans), len(set(source_scans)))
self.assertEqual(response.status_code, status.HTTP_200_OK)
coverage = response.data["data"]["registry_data_coverage"]