feat(provenance): govern source snapshots and data inputs

backend/tests/test_sprint107_map_selection_export.py

@@ -11,7 +11,7 @@ from shapely.geometry import box
 
 from app.core.errors import AppError
 from app.main import app
-from app.models import Area, Dataset, Export
+from app.models import Area, Dataset, Export, SourceRegistry, SourceSnapshot
 from app.schemas.export import ExportCreateResponse
 from app.services.export_service import ExportService
 from app.services.storage_service import StorageService
@@ -45,18 +45,56 @@ class FakeSession:
         return row
 
 
+def _govern_fixture_dataset(dataset: Dataset) -> Dataset:
+    source_id = uuid4()
+    snapshot_id = uuid4()
+    checksum = "a" * 64
+    source = SourceRegistry(
+        id=source_id,
+        source_key="grb",
+        display_name="GRB map export test source",
+        classification="authoritative",
+        authority_name="Digitaal Vlaanderen",
+        authority_scope_json={"zone": "Flanders"},
+        usage_policy_json={"ground_truth_allowed": True},
+    )
+    snapshot = SourceSnapshot(
+        id=snapshot_id,
+        source_registry_id=source_id,
+        snapshot_key=f"map-export-grb-{dataset.id}",
+        checksum_sha256=checksum,
+        ingest_status="ingested",
+        freshness_status="current",
+    )
+    dataset.source = "grb"
+    dataset.source_name = "grb"
+    dataset.checksum_sha256 = checksum
+    dataset.source_registry_id = source_id
+    dataset.source_snapshot_id = snapshot_id
+    dataset.data_contract_key = "geointel.vector.geojson"
+    dataset.data_contract_version = "1.0.0"
+    dataset.validation_status = "passed"
+    dataset.provenance_status = "complete"
+    dataset.lineage_status = "not_applicable"
+    dataset.quarantine_status = "not_quarantined"
+    dataset.status = "ready"
+    dataset.source_registry = source
+    dataset.source_snapshot = snapshot
+    return dataset
+
+
 def test_vector_selection_geojson_export_persists_handoff_artifact(tmp_path, monkeypatch) -> None:
     project_id = uuid4()
     dataset_id = uuid4()
     export_path = tmp_path / "exports" / "selection.geojson"
-    dataset = Dataset(
+    dataset = _govern_fixture_dataset(Dataset(
         id=dataset_id,
         project_id=project_id,
         name="candidate.geojson",
         dataset_type="vector",
         source="fixture",
         status="ready",
-    )
+    ))
     db = FakeSession({(Dataset, dataset_id): dataset})
     selection_bbox = {"min_x": 4.9, "min_y": 50.9, "max_x": 5.2, "max_y": 51.2, "crs": "EPSG:4326"}
     selection_payload = {
@@ -135,14 +173,14 @@ def test_vector_selection_export_uses_exact_area_scope_when_requested(tmp_path,
     project_id = uuid4()
     dataset_id = uuid4()
     area_id = uuid4()
-    dataset = Dataset(
+    dataset = _govern_fixture_dataset(Dataset(
         id=dataset_id,
         project_id=project_id,
         name="regional-buildings.geojson",
         dataset_type="vector",
         source="fixture",
         status="ready",
-    )
+    ))
     area_shape = box(5.0, 51.1, 5.2, 51.3)
     area_geometry = from_shape(area_shape, srid=4326)
     area = SimpleNamespace(id=area_id, project_id=project_id, name="Gemeente Mol", geometry=area_geometry)
@@ -184,7 +222,7 @@ def test_area_constrained_bbox_uses_intersection_and_disables_full_area_fast_pat
     project_id = uuid4()
     dataset_id = uuid4()
     area_id = uuid4()
-    dataset = Dataset(
+    dataset = _govern_fixture_dataset(Dataset(
         id=dataset_id,
         project_id=project_id,
         area_id=area_id,
@@ -193,7 +231,7 @@ def test_area_constrained_bbox_uses_intersection_and_disables_full_area_fast_pat
         source="fixture",
         source_metadata={"geometry_clipped_to_area": True},
         status="ready",
-    )
+    ))
     area_shape = box(5.0, 51.0, 5.2, 51.2)
     area = SimpleNamespace(
         id=area_id,