feast 0.66.0.dev69__tar.gz → 0.66.0.dev71__tar.gz

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/.secrets.baseline

@@ -989,7 +989,7 @@
         "filename": "infra/feast-operator/api/v1alpha1/featurestore_types.go",
         "hashed_secret": "44e17306b837162269a410204daaa5ecee4ec22c",
         "is_verified": false,
-        "line_number": 663
+        "line_number": 664
       }
     ],
     "infra/feast-operator/api/v1alpha1/zz_generated.deepcopy.go": [
@@ -1392,7 +1392,7 @@
         "filename": "sdk/python/tests/unit/infra/offline_stores/contrib/postgres_offline_store/test_postgres.py",
         "hashed_secret": "9fb7fe1217aed442b04c0f5e43b5d5a7d3287097",
         "is_verified": false,
-        "line_number": 301
+        "line_number": 364
       }
     ],
     "sdk/python/tests/unit/infra/offline_stores/test_clickhouse.py": [
@@ -1469,14 +1469,14 @@
         "filename": "sdk/python/tests/unit/permissions/test_oidc_auth_client.py",
         "hashed_secret": "e6eae2da3b4a5bf296d0495192788e2772ac5c79",
         "is_verified": false,
-        "line_number": 29
+        "line_number": 47
       },
       {
         "type": "Secret Keyword",
         "filename": "sdk/python/tests/unit/permissions/test_oidc_auth_client.py",
         "hashed_secret": "8318df9ecda039deac9868adf1944a29a95c7114",
         "is_verified": false,
-        "line_number": 31
+        "line_number": 49
       }
     ],
     "sdk/python/tests/universal/feature_repos/repo_configuration.py": [
@@ -1564,5 +1564,5 @@
       }
     ]
   },
-  "generated_at": "2026-07-31T18:15:46Z"
+  "generated_at": "2026-08-14T07:27:08Z"
 }

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: feast
-Version: 0.66.0.dev69
+Version: 0.66.0.dev71
 Summary: Python SDK for Feast
 Author: Feast
 License-Expression: Apache-2.0

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/pyproject.toml

@@ -324,7 +324,8 @@ python = "~=3.10.0"
 feast = { path = ".", editable = true, extras = ["duckdb", "delta", "grpcio", "test"] }
 
 [tool.pixi.feature.duckdb-tests.tasks]
-test = { cmd = "python -m pytest -n 8 --integration -m 'not ray_offline_stores_only' --ignore=sdk/python/tests/integration/offline_store/test_dqm_validation.py sdk/python/tests/integration/offline_store", env = { PYTHONPATH = ".", FULL_REPO_CONFIGS_MODULE = "sdk.python.tests.universal.feature_repos.duckdb_repo_configuration", FEAST_IS_LOCAL_TEST = "True" } }
+install-delta-extension = "python -c \"import duckdb; duckdb.connect().install_extension('delta')\""
+test = { depends-on = ["install-delta-extension"], cmd = "python -m pytest -n 8 --integration -m 'not ray_offline_stores_only' --ignore=sdk/python/tests/integration/offline_store/test_dqm_validation.py sdk/python/tests/integration/offline_store", env = { PYTHONPATH = ".", FULL_REPO_CONFIGS_MODULE = "sdk.python.tests.universal.feature_repos.duckdb_repo_configuration", FEAST_IS_LOCAL_TEST = "True" } }
 
 [tool.pixi.feature.ray-tests.pypi-dependencies]
 feast = { path = ".", editable = true, extras = ["ray", "grpcio", "test"] }

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/feature_store.py

@@ -3806,20 +3806,19 @@ class FeatureStore:
         Fails if the dataframe columns do not match the columns of the batch data source. Optionally
         reorders the columns of the dataframe to match.
         """
-        # TODO: restrict this to work with online StreamFeatureViews and validate the FeatureView type
-        try:
-            feature_view: FeatureView = self.get_stream_feature_view(
-                feature_view_name, allow_registry_cache=allow_registry_cache
-            )
-        except FeatureViewNotFoundException:
-            try:
-                feature_view = self.get_feature_view(
-                    feature_view_name, allow_registry_cache=allow_registry_cache
-                )
-            except FeatureViewNotFoundException:
-                feature_view = self.get_label_view(  # type: ignore[assignment]
-                    feature_view_name, allow_registry_cache=allow_registry_cache
-                )
+        # Resolve the feature view with a single registry lookup regardless of its
+        # type. The previous try/except chain tried get_stream_feature_view, then
+        # get_feature_view, then get_label_view in turn, so the common plain
+        # FeatureView always paid one guaranteed-miss lookup first. On a
+        # RemoteRegistry each miss is a wasted gRPC round-trip on this per-batch
+        # write path (see #6671).
+        # TODO: validate that the resolved feature view type supports offline writes.
+        feature_view = cast(
+            FeatureView,
+            self.registry.get_any_feature_view(
+                feature_view_name, self.project, allow_cache=allow_registry_cache
+            ),
+        )
 
         provider = self._get_provider()
         # Get columns of the batch source and the input dataframe.

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/infra/compute_engines/spark/feature_builder.py

@@ -109,7 +109,10 @@ class SparkFeatureBuilder(FeatureBuilder):
     def build_transformation_node(self, view, input_nodes):
         udf_name = view.feature_transformation.name
         udf = view.feature_transformation.udf
-        node = SparkTransformationNode(udf_name, udf, inputs=input_nodes)
+        udf_string = getattr(view.feature_transformation, "udf_string", "") or ""
+        node = SparkTransformationNode(
+            udf_name, udf, inputs=input_nodes, udf_string=udf_string
+        )
         self.nodes.append(node)
         return node
 

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/infra/compute_engines/spark/nodes.py

@@ -45,6 +45,7 @@ from feast.infra.offline_stores.contrib.spark_offline_store.spark import (
 from feast.infra.offline_stores.contrib.spark_offline_store.spark_source import (
     SparkSource,
 )
+from feast.transformation.udf_rehydrate import resolve_udf
 
 logger = logging.getLogger(__name__)
 
@@ -602,9 +603,32 @@ class SparkWriteNode(DAGNode):
 
 
 class SparkTransformationNode(DAGNode):
-    def __init__(self, name: str, udf: Callable, inputs: List[DAGNode]):
+    def __init__(
+        self,
+        name: str,
+        udf: Callable,
+        inputs: List[DAGNode],
+        udf_string: str = "",
+    ):
         super().__init__(name, inputs)
         self.udf = udf
+        self.udf_string = udf_string or ""
+        self._resolved_udf: Optional[Callable] = None
+
+    def _resolve_udf(self) -> Callable:
+        """Prefer source reconstruction over dill callables.
+
+        Dill-deserialized functions that call DataFrame.withColumn / __getitem__
+        can segfault (exit 139) on Spark 4.0.1. Re-executing ``udf_string``
+        yields a healthy callable. Result is cached for the lifetime of the node.
+        """
+        if self._resolved_udf is None:
+            self._resolved_udf = resolve_udf(
+                udf_string=self.udf_string,
+                fallback_udf=self.udf,
+                preferred_name=self.name,
+            )
+        return self._resolved_udf
 
     def execute(self, context: ExecutionContext) -> DAGValue:
         input_values = self.get_input_values(context)
@@ -613,7 +637,7 @@ class SparkTransformationNode(DAGNode):
 
         input_dfs: List[DataFrame] = [val.data for val in input_values]
 
-        transformed_df = self.udf(*input_dfs)
+        transformed_df = self._resolve_udf()(*input_dfs)
 
         return DAGValue(
             data=transformed_df, format=DAGFormat.SPARK, metadata={"transformed": True}

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/infra/offline_stores/contrib/postgres_offline_store/postgres.py

@@ -101,14 +101,35 @@ class PostgreSQLOfflineStore(OfflineStore):
         if created_timestamp_column:
             timestamps.append(created_timestamp_column)
         timestamp_desc_string = " DESC, ".join(_append_alias(timestamps, "a")) + " DESC"
-        a_field_string = ", ".join(
-            _append_alias(join_key_columns + feature_name_columns + timestamps, "a")
-        )
-        b_field_string = ", ".join(
-            _append_alias(join_key_columns + feature_name_columns + timestamps, "b")
-        )
-
-        query = f"""
+        # Empty feature_name_columns means "all source columns". BatchFeatureView
+        # python/pandas/ray transforms signal this via get_column_info. Selecting
+        # only join keys + timestamps would starve the UDF of input features.
+        if not feature_name_columns:
+            distinct_on = ", ".join(f'a."{c}"' for c in join_key_columns) or (
+                f'a."{timestamp_field}"'
+            )
+            order_by_parts = [f'a."{c}"' for c in join_key_columns] + [
+                f'a."{timestamp_field}" DESC'
+            ]
+            if created_timestamp_column:
+                order_by_parts.append(f'a."{created_timestamp_column}" DESC')
+            order_by = ", ".join(order_by_parts)
+            query = f"""
+            SELECT DISTINCT ON ({distinct_on})
+                a.*
+                {f", {repr(DUMMY_ENTITY_VAL)} AS {DUMMY_ENTITY_ID}" if not join_key_columns else ""}
+            FROM {from_expression} a
+            WHERE a."{timestamp_field}" BETWEEN '{start_date}'::timestamptz AND '{end_date}'::timestamptz
+            ORDER BY {order_by}
+            """
+        else:
+            a_field_string = ", ".join(
+                _append_alias(join_key_columns + feature_name_columns + timestamps, "a")
+            )
+            b_field_string = ", ".join(
+                _append_alias(join_key_columns + feature_name_columns + timestamps, "b")
+            )
+            query = f"""
             SELECT
                 {b_field_string}
                 {f", {repr(DUMMY_ENTITY_VAL)} AS {DUMMY_ENTITY_ID}" if not join_key_columns else ""}
@@ -275,12 +296,17 @@ class PostgreSQLOfflineStore(OfflineStore):
         timestamp_fields = [timestamp_field]
         if created_timestamp_column:
             timestamp_fields.append(created_timestamp_column)
-        field_string = ", ".join(
-            _append_alias(
-                join_key_columns + feature_name_columns + timestamp_fields,
-                "paftoq_alias",
+        # Empty feature_name_columns => SELECT * (BatchFeatureView python mode).
+        # Default materialization uses pull_all (pull_latest_features=False).
+        if not feature_name_columns:
+            field_string = "paftoq_alias.*"
+        else:
+            field_string = ", ".join(
+                _append_alias(
+                    join_key_columns + feature_name_columns + timestamp_fields,
+                    "paftoq_alias",
+                )
             )
-        )
 
         timestamp_filter = get_timestamp_filter_sql(
             start_date,

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/infra/offline_stores/contrib/trino_offline_store/trino_type_map.py

@@ -22,6 +22,9 @@ def trino_to_feast_value_type(trino_type_as_str: str) -> ValueType:
         "boolean": ValueType.BOOL,
         "real": ValueType.FLOAT,
         "date": ValueType.STRING,
+        "binary": ValueType.STRING,
+        "varbinary": ValueType.STRING,
+        "json": ValueType.STRING,
     }
     _trino_type_as_str: str = trino_type_as_str
     trino_type_as_str = trino_type_as_str.lower()
@@ -36,6 +39,8 @@ def trino_to_feast_value_type(trino_type_as_str: str) -> ValueType:
                 trino_type_as_str = "decimal64"
             else:
                 trino_type_as_str = "decimal32"
+        else:
+            trino_type_as_str = "decimal64"
 
     elif trino_type_as_str.startswith("timestamp"):
         trino_type_as_str = "timestamp"
@@ -43,6 +48,9 @@ def trino_to_feast_value_type(trino_type_as_str: str) -> ValueType:
     elif trino_type_as_str.startswith("varchar"):
         trino_type_as_str = "varchar"
 
+    elif trino_type_as_str.startswith("char"):
+        trino_type_as_str = "char"
+
     if trino_type_as_str not in type_map:
         raise ValueError(f"Trino type not supported by feast {_trino_type_as_str}")
     return type_map[trino_type_as_str]
@@ -55,7 +63,11 @@ def pa_to_trino_value_type(pa_type_as_str: str) -> str:
     trino_type = "{}"
     if pa_type_as_str.startswith("list"):
         trino_type = "array<{}>"
-        pa_type_as_str = re.search(r"^list<item:\s(.+)>$", pa_type_as_str).group(1)
+        match = re.search(r"^list<item:\s(.+)>$", pa_type_as_str)
+        if match:
+            pa_type_as_str = match.group(1)
+        else:
+            return trino_type.format("varchar")
 
     if pa_type_as_str.startswith("date"):
         return trino_type.format("date")
@@ -67,7 +79,10 @@ def pa_to_trino_value_type(pa_type_as_str: str) -> str:
             return trino_type.format("timestamp")
 
     if pa_type_as_str.startswith("decimal"):
-        return trino_type.format(pa_type_as_str)
+        # PyArrow renders decimal types as decimal128(10, 2) or decimal256(10, 2),
+        # but Trino expects just decimal(10, 2)
+        normalized = re.sub(r"^decimal\d+", "decimal", pa_type_as_str)
+        return trino_type.format(normalized)
 
     if pa_type_as_str.startswith("map<"):
         return trino_type.format("varchar")
@@ -92,33 +107,43 @@ def pa_to_trino_value_type(pa_type_as_str: str) -> str:
         "float": "double",
         "double": "double",
         "binary": "binary",
+        "varbinary": "binary",
         "string": "varchar",
+        "char": "varchar",
     }
     return trino_type.format(type_map[pa_type_as_str])
 
 
-_TRINO_TO_PA_TYPE_MAP = {
+_TRINO_TO_PA_TYPE_MAP: Dict[str, pa.DataType] = {
     "null": pa.null(),
     "boolean": pa.bool_(),
     "date": pa.date32(),
     "tinyint": pa.int8(),
     "smallint": pa.int16(),
     "integer": pa.int32(),
+    "int": pa.int32(),
     "bigint": pa.int64(),
     "double": pa.float64(),
     "binary": pa.binary(),
+    "varbinary": pa.binary(),
     "char": pa.string(),
+    "json": pa.string(),
     "real": pa.float32(),
 }
 
 
+def _trino_array_item_type(trino_type_as_str: str) -> str | None:
+    if trino_type_as_str.startswith("array(") and trino_type_as_str.endswith(")"):
+        return trino_type_as_str[6:-1].strip()
+    return None
+
+
 def trino_to_pa_value_type(trino_type_as_str: str) -> pa.DataType:
-    trino_type_as_str = trino_type_as_str.lower()
+    trino_type_as_str = trino_type_as_str.lower().strip()
 
-    _is_list: bool = False
-    if trino_type_as_str.startswith("array"):
-        _is_list = True
-        trino_type_as_str = re.search(r"^array\((\w+)\)$", trino_type_as_str).group(1)
+    array_item_type = _trino_array_item_type(trino_type_as_str)
+    if array_item_type is not None:
+        return pa.list_(trino_to_pa_value_type(array_item_type))
 
     if trino_type_as_str.startswith("decimal"):
         search_precision = re.search(
@@ -127,20 +152,24 @@ def trino_to_pa_value_type(trino_type_as_str: str) -> pa.DataType:
         if search_precision:
             precision = int(search_precision.group(1))
             if precision > 32:
-                pa_type = pa.float64()
+                return pa.float64()
             else:
-                pa_type = pa.float32()
+                return pa.float32()
+        return pa.float64()
 
-    elif trino_type_as_str.startswith("timestamp"):
-        pa_type = pa.timestamp("us")
+    if trino_type_as_str.startswith("timestamp"):
+        return pa.timestamp("us")
 
-    elif trino_type_as_str.startswith("varchar"):
-        pa_type = pa.string()
+    if trino_type_as_str.startswith("varchar"):
+        return pa.string()
+
+    if trino_type_as_str.startswith("char"):
+        return pa.string()
+
+    if trino_type_as_str.startswith("row("):
+        return pa.string()
 
-    else:
-        pa_type = _TRINO_TO_PA_TYPE_MAP[trino_type_as_str]
+    if trino_type_as_str.startswith("map("):
+        return pa.string()
 
-    if _is_list:
-        return pa.list_(pa_type)
-    else:
-        return pa_type
+    return _TRINO_TO_PA_TYPE_MAP[trino_type_as_str]

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/infra/registry/remote.py

@@ -489,7 +489,7 @@ class RemoteRegistry(BaseRegistry):
         request = RegistryServer_pb2.ApplySavedDatasetRequest(
             saved_dataset=saved_dataset.to_proto(), project=project, commit=commit
         )
-        self.stub.ApplyFeatureService(request)
+        self.stub.ApplySavedDataset(request)
 
     def delete_saved_dataset(self, name: str, project: str, commit: bool = True):
         request = RegistryServer_pb2.DeleteSavedDatasetRequest(

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/permissions/auth_model.py

@@ -67,6 +67,11 @@ class OidcClientAuthConfig(OidcAuthConfig):
     client_secret: Optional[str] = None
     token: Optional[str] = None
     token_env_var: Optional[str] = None
+    # Stop reusing an IdP-issued token this many seconds before it expires,
+    # so a reused token still has life left when the server validates it.
+    # Raise it if clients see sporadic 401s from clock skew or slow calls;
+    # lower it to squeeze more reuse out of short-lived tokens.
+    token_refresh_margin_seconds: float = Field(default=30, gt=0)
 
     @model_validator(mode="after")
     def _validate_credentials(self):

feast-0.66.0.dev71/sdk/python/feast/permissions/client/client_auth_token.py

@@ -0,0 +1,27 @@
+from feast.permissions.auth_model import (
+    AuthConfig,
+)
+from feast.permissions.client.auth_client_manager import (
+    AuthenticationClientManagerFactory,
+)
+
+
+def get_auth_token(auth_config: AuthConfig) -> str:
+    return (
+        AuthenticationClientManagerFactory(auth_config)
+        .get_auth_client_manager()
+        .get_token()
+    )
+
+
+def invalidate_auth_token(auth_config: AuthConfig) -> bool:
+    """Drop any cached token for *auth_config*, returning whether one was held.
+
+    Only the OIDC client manager caches, so this is a no-op for the other auth
+    types. Callers that can observe an authentication failure should use it: a
+    token the IdP revokes mid-life still looks valid to the client until its
+    own expiry, and dropping it bounds that to a single rejected request.
+    """
+    manager = AuthenticationClientManagerFactory(auth_config).get_auth_client_manager()
+    invalidate = getattr(manager, "invalidate_token", None)
+    return bool(invalidate()) if callable(invalidate) else False

{feast-0.66.0.dev69 → feast-0.66.0.dev71}/sdk/python/feast/permissions/client/grpc_client_auth_interceptor.py

@@ -5,7 +5,10 @@ import grpc
 from feast.errors import FeastError
 from feast.permissions.auth.auth_type import AuthType
 from feast.permissions.auth_model import AuthConfig
-from feast.permissions.client.client_auth_token import get_auth_token
+from feast.permissions.client.client_auth_token import (
+    get_auth_token,
+    invalidate_auth_token,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -44,11 +47,37 @@ class GrpcClientAuthHeaderInterceptor(
             client_call_details = self._append_auth_header_metadata(client_call_details)
         result = continuation(client_call_details, request_iterator)
         if result.exception() is not None:
+            self._invalidate_token_if_rejected(result)
             mapped_error = FeastError.from_error_detail(result.exception().details())
             if mapped_error is not None:
                 raise mapped_error
         return result
 
+    def _invalidate_token_if_rejected(self, result) -> None:
+        """Drop the cached token when the server rejects it as unauthenticated.
+
+        Tokens are reused until near expiry, so one the IdP revoked mid-life
+        would otherwise keep being presented for the rest of its lifetime.
+        Dropping it here bounds that to the single request that was rejected;
+        the next call fetches a fresh token.
+
+        The call is deliberately not retried. All four interceptor methods
+        share this path, and a stream's ``request_iterator`` may already be
+        consumed, so retrying here could replay a partially-sent stream.
+        """
+        if self._auth_config.type == AuthType.NONE.value:
+            return
+        try:
+            if result.code() != grpc.StatusCode.UNAUTHENTICATED:
+                return
+        except Exception:  # pragma: no cover - result without a status code
+            return
+        if invalidate_auth_token(self._auth_config):
+            logger.debug(
+                "Server rejected the cached auth token; dropped it so the next "
+                "call fetches a fresh one."
+            )
+
     def _append_auth_header_metadata(self, client_call_details):
         logger.debug(
             "Intercepted the grpc api method call to inject Authorization header "