tencent-wedata-feature-engineering-dev 0.1.44__tar.gz → 0.1.45__tar.gz

{tencent-wedata-feature-engineering-dev-0.1.44 → tencent-wedata-feature-engineering-dev-0.1.45}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: tencent-wedata-feature-engineering-dev
-Version: 0.1.44
+Version: 0.1.45
 Summary: Wedata Feature Engineering Library Development
 Home-page: 
 Author: meahqian

{tencent-wedata-feature-engineering-dev-0.1.44 → tencent-wedata-feature-engineering-dev-0.1.45}/tencent_wedata_feature_engineering_dev.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: tencent-wedata-feature-engineering-dev
-Version: 0.1.44
+Version: 0.1.45
 Summary: Wedata Feature Engineering Library Development
 Home-page: 
 Author: meahqian

{tencent-wedata-feature-engineering-dev-0.1.44 → tencent-wedata-feature-engineering-dev-0.1.45}/wedata/feature_store/constants/constants.py

@@ -43,7 +43,7 @@ ML_MODEL = "MLmodel"
 FEATURE_LOOKUP_CLIENT_PIP_PACKAGE = "tencent-wedata-feature-engineering-dev"
 
 # 特征查找版本号
-FEATURE_LOOKUP_CLIENT_MAJOR_VERSION = "0.1.44"
+FEATURE_LOOKUP_CLIENT_MAJOR_VERSION = "0.1.45"
 
 # 特征存储内部数据目录
 FEATURE_STORE_INTERNAL_DATA_DIR = "_wedata_internal/"

{tencent-wedata-feature-engineering-dev-0.1.44 → tencent-wedata-feature-engineering-dev-0.1.45}/wedata/feature_store/feature_table_client/feature_table_client.py

@@ -123,6 +123,11 @@ class FeatureTableClient:
         """转义SQL值中的特殊字符"""
         return value.replace("'", "''")
 
+    @staticmethod
+    def _check_sequence_element_type(sequence: Sequence[Any], element_type: type) -> bool:
+        """检查序列中的元素是否为指定类型"""
+        return all(isinstance(element, element_type) for element in sequence)
+
     def create_table(
             self,
             name: str,
@@ -165,6 +170,10 @@ class FeatureTableClient:
         primary_keys = self._normalize_params(primary_keys)
         partition_columns = self._normalize_params(partition_columns)
 
+        assert self._check_sequence_element_type(primary_keys, str), "primary_keys must be a list of strings"
+        assert self._check_sequence_element_type(partition_columns, str), "partition_columns must be a list of strings"
+        assert isinstance(timestamp_key, str), "timestamp key must be string"
+
         # 元数据校验
         self._validate_schema(df, schema)
         self._validate_key_exists(primary_keys, timestamp_key)
@@ -444,6 +453,10 @@ class FeatureTableClient:
         common_utils.validate_table_name(name)
         common_utils.validate_database(database_name)
 
+        if primary_keys:
+            assert self._check_sequence_element_type(primary_keys, str), "primary_keys must be a list of strings"
+        assert isinstance(timestamp_key, str), "timestamp key must be string"
+
         # 构建完整表名
         table_name = common_utils.build_full_table_name(name, database_name)
 
@@ -456,11 +469,12 @@ class FeatureTableClient:
 
             # 检查Primary Key和Timestamp Key是否为空
             if engine_type == engine_type.ICEBERG_ENGINE and props.get("format-version", "") == "2":
-                if props.get('dlc.ao.data.govern.sorted.keys', "") == "":
-                    raise ValueError(
-                        "table dlc.ao.data.govern.sorted.keys is empty. you must set dlc.ao.data.govern.sorted.keys")
-                else:
-                    primary_keys = props.get('dlc.ao.data.govern.sorted.keys').split(",")
+                if not primary_keys:
+                    if props.get('dlc.ao.data.govern.sorted.keys', "") == "":
+                        raise ValueError(
+                            "table dlc.ao.data.govern.sorted.keys is empty. you must set dlc.ao.data.govern.sorted.keys")
+                    else:
+                        primary_keys = props.get('dlc.ao.data.govern.sorted.keys').split(",")
             elif engine_type == engine_type.HIVE_ENGINE:
                 if not primary_keys:
                     raise ValueError("primary_keys cannot be None for HIVE_ENGINE")
@@ -613,9 +627,6 @@ class FeatureTableClient:
                 print(f"Table '{name}' does not exist")
                 return
 
-            # 检查
-            df = self._spark.table(tableName=table_name)
-
             try:
                 feature_view = self._feast_client.get_feature_view(table_name)
             except Exception as e: