fix: Infer features for feature services when they depend on feature …

…views without schemas (#2653)

* fix: Infer features for feature services when they depend on feature views that have no schema

Signed-off-by: Danny Chiao <danny@tecton.ai>

* fix

Signed-off-by: Danny Chiao <danny@tecton.ai>

* lint

Signed-off-by: Danny Chiao <danny@tecton.ai>

sdk/python/feast/feature_service.py

@@ -39,6 +39,7 @@ class FeatureService:
     """
 
     name: str
+    _features: List[Union[FeatureView, OnDemandFeatureView]]
     feature_view_projections: List[FeatureViewProjection]
     description: str
     tags: Dict[str, str]
@@ -93,24 +94,36 @@ def __init__(
             _features = []
 
         self.name = _name
+        self._features = _features
         self.feature_view_projections = []
+        self.description = description
+        self.tags = tags or {}
+        self.owner = owner
+        self.created_timestamp = None
+        self.last_updated_timestamp = None
+        self.logging_config = logging_config
+        self.infer_features()
 
-        for feature_grouping in _features:
+    def infer_features(self, fvs_to_update: Optional[Dict[str, FeatureView]] = None):
+        self.feature_view_projections = []
+        for feature_grouping in self._features:
             if isinstance(feature_grouping, BaseFeatureView):
+                # For feature services that depend on an unspecified feature view, apply inferred schema
+                if (
+                    fvs_to_update is not None
+                    and len(feature_grouping.projection.features) == 0
+                    and feature_grouping.name in fvs_to_update
+                ):
+                    feature_grouping.projection.features = fvs_to_update[
+                        feature_grouping.name
+                    ].features
                 self.feature_view_projections.append(feature_grouping.projection)
             else:
                 raise ValueError(
-                    f"The feature service {name} has been provided with an invalid type "
+                    f"The feature service {self.name} has been provided with an invalid type "
                     f'{type(feature_grouping)} as part of the "features" argument.)'
                 )
 
-        self.description = description
-        self.tags = tags or {}
-        self.owner = owner
-        self.created_timestamp = None
-        self.last_updated_timestamp = None
-        self.logging_config = logging_config
-
     def __repr__(self):
         items = (f"{k} = {v}" for k, v in self.__dict__.items())
         return f"<{self.__class__.__name__}({', '.join(items)})>"
@@ -119,7 +132,7 @@ def __str__(self):
         return str(MessageToJson(self.to_proto()))
 
     def __hash__(self):
-        return hash((self.name))
+        return hash(self.name)
 
     def __eq__(self, other):
         if not isinstance(other, FeatureService):

sdk/python/feast/feature_store.py

@@ -475,8 +475,9 @@ def _make_inferences(
         entities_to_update: List[Entity],
         views_to_update: List[FeatureView],
         odfvs_to_update: List[OnDemandFeatureView],
+        feature_services_to_update: List[FeatureService],
     ):
-        """Makes inferences for entities, feature views, and odfvs."""
+        """Makes inferences for entities, feature views, odfvs, and feature services."""
         update_entities_with_inferred_types_from_feature_views(
             entities_to_update, views_to_update, self.config
         )
@@ -498,6 +499,10 @@ def _make_inferences(
         for odfv in odfvs_to_update:
             odfv.infer_features()
 
+        fvs_to_update_map = {view.name: view for view in views_to_update}
+        for feature_service in feature_services_to_update:
+            feature_service.infer_features(fvs_to_update=fvs_to_update_map)
+
     @log_exceptions_and_usage
     def _plan(
         self, desired_repo_contents: RepoContents
@@ -553,6 +558,7 @@ def _plan(
             desired_repo_contents.entities,
             desired_repo_contents.feature_views,
             desired_repo_contents.on_demand_feature_views,
+            desired_repo_contents.feature_services,
         )
 
         # Compute the desired difference between the current objects in the registry and
@@ -692,7 +698,11 @@ def apply(
             views_to_update, odfvs_to_update, request_views_to_update
         )
         self._make_inferences(
-            data_sources_to_update, entities_to_update, views_to_update, odfvs_to_update
+            data_sources_to_update,
+            entities_to_update,
+            views_to_update,
+            odfvs_to_update,
+            services_to_update,
         )
 
         # Handle all entityless feature views by using DUMMY_ENTITY as a placeholder entity.

sdk/python/tests/integration/registration/test_inference.py

@@ -7,6 +7,7 @@
     BigQuerySource,
     Entity,
     Feature,
+    FeatureService,
     FileSource,
     RedshiftSource,
     RepoConfig,
@@ -332,3 +333,25 @@ def test_update_feature_views_with_inferred_features():
     )
     assert len(feature_view_2.schema) == 1
     assert len(feature_view_2.features) == 1
+
+
+def test_update_feature_services_with_inferred_features(simple_dataset_1):
+    with prep_file_source(df=simple_dataset_1, timestamp_field="ts_1") as file_source:
+        entity1 = Entity(name="test1", join_keys=["id_join_key"])
+        feature_view_1 = FeatureView(
+            name="test1", entities=[entity1], source=file_source,
+        )
+        feature_service = FeatureService(name="fs_1", features=[feature_view_1])
+        assert len(feature_service.feature_view_projections) == 1
+        assert len(feature_service.feature_view_projections[0].features) == 0
+
+        update_feature_views_with_inferred_features(
+            [feature_view_1], [entity1], RepoConfig(provider="local", project="test")
+        )
+        feature_service.infer_features(
+            fvs_to_update={feature_view_1.name: feature_view_1}
+        )
+
+        assert len(feature_view_1.schema) == 3
+        assert len(feature_view_1.features) == 3
+        assert len(feature_service.feature_view_projections[0].features) == 3