feat: harden datasource materialization payloads
This commit is contained in:
@@ -17,11 +17,14 @@ from govoplan_core.core.datasources import (
|
||||
DatasourcePublicationRequest,
|
||||
DatasourceReadRequest,
|
||||
DatasourceStageInput,
|
||||
DatasourceUnavailableError,
|
||||
DatasourceValidationError,
|
||||
)
|
||||
from govoplan_core.db.base import Base, utcnow
|
||||
from govoplan_datasources.backend.db.models import (
|
||||
DatasourceMaterializationRecord,
|
||||
DatasourcePayloadRecord,
|
||||
DatasourcePayloadRowRecord,
|
||||
DatasourcePublicationRecord,
|
||||
DatasourceRecord,
|
||||
DatasourceStageRecord,
|
||||
@@ -32,6 +35,12 @@ from govoplan_datasources.backend.service import (
|
||||
STAGE_WRITE_SCOPE,
|
||||
SqlDatasourceProvider,
|
||||
)
|
||||
from govoplan_datasources.backend.payloads import (
|
||||
create_database_rows_payload,
|
||||
finalize_payload_deletion,
|
||||
mark_unreferenced_payload_for_deletion,
|
||||
verify_payload_integrity,
|
||||
)
|
||||
|
||||
|
||||
def principal(
|
||||
@@ -137,6 +146,8 @@ class DatasourceLifecycleTests(unittest.TestCase):
|
||||
self.engine,
|
||||
tables=[
|
||||
DatasourceRecord.__table__,
|
||||
DatasourcePayloadRecord.__table__,
|
||||
DatasourcePayloadRowRecord.__table__,
|
||||
DatasourceMaterializationRecord.__table__,
|
||||
DatasourceStageRecord.__table__,
|
||||
DatasourcePublicationRecord.__table__,
|
||||
@@ -157,6 +168,8 @@ class DatasourceLifecycleTests(unittest.TestCase):
|
||||
DatasourceStageRecord.__table__,
|
||||
DatasourcePublicationRecord.__table__,
|
||||
DatasourceMaterializationRecord.__table__,
|
||||
DatasourcePayloadRowRecord.__table__,
|
||||
DatasourcePayloadRecord.__table__,
|
||||
DatasourceRecord.__table__,
|
||||
],
|
||||
)
|
||||
@@ -235,6 +248,18 @@ class DatasourceLifecycleTests(unittest.TestCase):
|
||||
current.datasource.schema_version,
|
||||
)
|
||||
self.assertEqual(frozen.ref, frozen_result.materialization.ref)
|
||||
first_record = self.session.get(
|
||||
DatasourceMaterializationRecord,
|
||||
first.ref.removeprefix("materialization:"),
|
||||
)
|
||||
frozen_record = self.session.get(
|
||||
DatasourceMaterializationRecord,
|
||||
frozen.ref.removeprefix("materialization:"),
|
||||
)
|
||||
self.assertIsNotNone(first_record)
|
||||
self.assertIsNotNone(frozen_record)
|
||||
self.assertEqual(first_record.payload_id, frozen_record.payload_id)
|
||||
self.assertEqual([], first_record.rows)
|
||||
|
||||
def test_live_reads_origin_and_cached_refresh_is_explicit(self) -> None:
|
||||
live = self.provider.register_origin(
|
||||
@@ -445,6 +470,114 @@ class DatasourceLifecycleTests(unittest.TestCase):
|
||||
),
|
||||
)
|
||||
|
||||
def test_payload_preview_is_paged_and_metadata_mismatch_is_rejected(self) -> None:
|
||||
stage = self.provider.create_stage(
|
||||
self.session,
|
||||
principal(),
|
||||
stage=DatasourceStageInput(
|
||||
name="Paged",
|
||||
source_name="paged",
|
||||
kind="upload",
|
||||
mode="static",
|
||||
shape="tabular",
|
||||
rows=tuple({"id": index} for index in range(20)),
|
||||
),
|
||||
)
|
||||
datasource, materialization = self.provider.promote_stage(
|
||||
self.session,
|
||||
principal(),
|
||||
stage_ref=stage.ref,
|
||||
)
|
||||
self.session.commit()
|
||||
|
||||
result = self.provider.read_datasource(
|
||||
self.session,
|
||||
principal(),
|
||||
request=DatasourceReadRequest(
|
||||
datasource_ref=datasource.ref,
|
||||
offset=7,
|
||||
limit=3,
|
||||
),
|
||||
)
|
||||
self.assertEqual([{"id": 7}, {"id": 8}, {"id": 9}], list(result.rows))
|
||||
record = self.session.get(
|
||||
DatasourceMaterializationRecord,
|
||||
materialization.ref.removeprefix("materialization:"),
|
||||
)
|
||||
self.assertIsNotNone(record)
|
||||
self.assertEqual([], record.rows)
|
||||
self.assertEqual(
|
||||
20,
|
||||
self.session.query(DatasourcePayloadRowRecord)
|
||||
.filter(DatasourcePayloadRowRecord.payload_id == record.payload_id)
|
||||
.count(),
|
||||
)
|
||||
|
||||
payload = self.session.get(DatasourcePayloadRecord, record.payload_id)
|
||||
self.assertIsNotNone(payload)
|
||||
verify_payload_integrity(self.session, payload)
|
||||
payload.row_count += 1
|
||||
self.session.flush()
|
||||
with self.assertRaises(DatasourceUnavailableError):
|
||||
self.provider.read_datasource(
|
||||
self.session,
|
||||
principal(),
|
||||
request=DatasourceReadRequest(datasource_ref=datasource.ref),
|
||||
)
|
||||
|
||||
def test_payload_deletion_is_staged_and_reference_safe(self) -> None:
|
||||
payload = create_database_rows_payload(
|
||||
self.session,
|
||||
tenant_id="tenant-1",
|
||||
rows=({"id": 1}, {"id": 2}),
|
||||
actor_id="account-1",
|
||||
)
|
||||
self.session.commit()
|
||||
|
||||
self.assertTrue(
|
||||
mark_unreferenced_payload_for_deletion(self.session, payload)
|
||||
)
|
||||
self.assertEqual("deleting", payload.state)
|
||||
self.session.commit()
|
||||
|
||||
finalize_payload_deletion(self.session, payload)
|
||||
self.session.commit()
|
||||
self.assertEqual(0, self.session.query(DatasourcePayloadRecord).count())
|
||||
self.assertEqual(
|
||||
0,
|
||||
self.session.query(DatasourcePayloadRowRecord).count(),
|
||||
)
|
||||
|
||||
def test_rolled_back_materialization_leaves_no_payload_rows(self) -> None:
|
||||
stage = self.provider.create_stage(
|
||||
self.session,
|
||||
principal(),
|
||||
stage=DatasourceStageInput(
|
||||
name="Rollback",
|
||||
source_name="rollback",
|
||||
kind="upload",
|
||||
mode="static",
|
||||
shape="tabular",
|
||||
rows=({"id": 1},),
|
||||
),
|
||||
)
|
||||
self.provider.promote_stage(
|
||||
self.session,
|
||||
principal(),
|
||||
stage_ref=stage.ref,
|
||||
)
|
||||
self.session.rollback()
|
||||
|
||||
self.assertEqual(0, self.session.query(DatasourcePayloadRecord).count())
|
||||
self.assertEqual(
|
||||
0,
|
||||
self.session.query(DatasourcePayloadRowRecord).count(),
|
||||
)
|
||||
self.assertEqual(
|
||||
0,
|
||||
self.session.query(DatasourceMaterializationRecord).count(),
|
||||
)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
unittest.main()
|
||||
|
||||
Reference in New Issue
Block a user