fix(platform): govern geospatial analysis and raster handoffs

backend/tests/test_segmentation_configured_models.py

@@ -9,11 +9,12 @@ import pytest
 from geoalchemy2.shape import to_shape
 
 from app.core.config import Settings
-from app.models import Dataset, Project, Segmentation, SourceRegistry, SourceSnapshot
+from app.models import Dataset, DatasetVersion, Project, Segmentation, SourceRegistry, SourceSnapshot
 from app.services.detection_georeferencing import pixel_points_to_epsg4326_polygon
 from app.services.model_registry_service import ModelRegistryService
 from app.services.runtime_model_provenance_service import RuntimeModelProvenanceService
 from app.services.segmentation_service import SegmentationService
+from app.services.tile_manifest_service import TileManifestService
 
 ROOT = Path(__file__).resolve().parents[2]
 
@@ -122,6 +123,8 @@ def _project_and_dataset(dataset_type: str = "raster"):
         source_name="digitaal_vlaanderen_orthophoto",
         storage_path="storage/uploads/ortho.tif",
         checksum_sha256=checksum,
+        crs="EPSG:4326",
+        bounds_json={"min_x": 4.0, "min_y": 51.0, "max_x": 5.0, "max_y": 52.0},
         source_registry_id=source_id,
         source_snapshot_id=snapshot_id,
         data_contract_key="geointel.raster.geotiff",
@@ -134,6 +137,9 @@ def _project_and_dataset(dataset_type: str = "raster"):
     )
     dataset.source_registry = source
     dataset.source_snapshot = snapshot
+    dataset.versions.append(
+        DatasetVersion(id=uuid4(), dataset_id=dataset_id, version=1, checksum_sha256=checksum)
+    )
     db = FakeSession(objects={(Project, project_id): project, (Dataset, dataset_id): dataset})
     return db, project_id, dataset_id
 
@@ -249,28 +255,36 @@ def _write_configured_model_sidecars(
         )
 
 
-def _manifest(tmp_path: Path, tile_count: int = 1) -> Path:
+def _manifest(
+    tmp_path: Path,
+    tile_count: int = 1,
+    *,
+    db: FakeSession | None = None,
+    dataset: Dataset | None = None,
+) -> Path:
     tiles = []
     for index in range(tile_count):
         tile_path = tmp_path / f"tile_{index:04d}.tif"
         tile_path.write_bytes(b"fixture")
-        tiles.append(
-            {
+        tile = {
                 "path": str(tile_path),
                 "pixel_window": [0, 0, 100, 100],
                 "bounds": [4.0, 51.0, 5.0, 52.0],
                 "transform": [4.0, 0.01, 0.0, 52.0, 0.0, -0.01],
                 "crs": "EPSG:4326",
                 "index": index,
+                **TileManifestService.tile_integrity(tile_path),
             }
-        )
+        tiles.append(tile)
+    binding = TileManifestService.dataset_binding(db or FakeSession(), dataset) if dataset is not None else {}
     manifest_path = tmp_path / "manifest.json"
     manifest_path.write_text(
         json.dumps(
             {
+                **binding,
                 "tile_set_id": "tiles-fixture",
-                "source_dataset_id": str(uuid4()),
-                "source_raster_id": str(uuid4()),
+                "source_dataset_id": binding.get("source_dataset_id", str(uuid4())),
+                "source_raster_id": binding.get("source_raster_id", str(uuid4())),
                 "crs": "EPSG:4326",
                 "bounds": [4.0, 51.0, 5.0, 52.0],
                 "tile_size": 100,
@@ -424,7 +438,9 @@ def test_configured_segmentation_rejects_unbound_model_snapshot_before_adapter_l
         dataset_id=dataset_id,
         model_id="yolo-seg-configured",
         confidence_threshold=0.5,
-        tile_manifest_path=str(_manifest(tmp_path)),
+        tile_manifest_path=str(
+            _manifest(tmp_path, db=db, dataset=db.get(Dataset, dataset_id))
+        ),
         settings=settings,
         yolo_seg_adapter_class=NeverLoadSegAdapter,
         sam_adapter_class=ClassAgnosticSamAdapter,
@@ -440,7 +456,7 @@ def test_configured_yolo_seg_run_persists_georeferenced_masks(tmp_path: Path) ->
     db, project_id, dataset_id = _project_and_dataset()
     settings = _settings(tmp_path)
     _write_configured_model_sidecars(tmp_path, settings, include_sam=False, db=db)
-    manifest_path = _manifest(tmp_path)
+    manifest_path = _manifest(tmp_path, db=db, dataset=db.get(Dataset, dataset_id))
 
     response = SegmentationService.run_segmentation(
         db=db,
@@ -479,7 +495,7 @@ def test_configured_sam_run_is_class_agnostic(tmp_path: Path) -> None:
     db, project_id, dataset_id = _project_and_dataset()
     settings = _settings(tmp_path)
     _write_configured_model_sidecars(tmp_path, settings, include_yolo=False, db=db)
-    manifest_path = _manifest(tmp_path)
+    manifest_path = _manifest(tmp_path, db=db, dataset=db.get(Dataset, dataset_id))
 
     response = SegmentationService.run_segmentation(
         db=db,