pyobvector 0.2.12__py3-none-any.whl → 0.2.13__py3-none-any.whl

pyobvector/client/ob_vec_client.py

@@ -89,6 +89,13 @@ class ObVecClient:
                         message=ExceptionsMessage.ClusterVersionIsLow,
                     )
 
+    def refresh_metadata(self, tables: Optional[list[str]] = None):
+        """Reload metadata from the database."""
+        if tables is None:
+            self.metadata_obj.reflect(bind=self.engine, extend_existing=True)
+        else:
+            self.metadata_obj.reflect(bind=self.engine, only=tables, extend_existing=True)
+
     def _insert_partition_hint_for_query_sql(self, sql: str, partition_hint: str):
         from_index = sql.find("FROM")
         assert from_index != -1

pyobvector/schema/array.py

@@ -10,7 +10,6 @@ class ARRAY(UserDefinedType):
     """ARRAY data type definition with support for up to 6 levels of nesting."""
     cache_ok = True
     _string = String()
-    _max_nesting_level = 6
 
     def __init__(self, item_type: Union[TypeEngine, type]):
         """Construct an ARRAY.
@@ -18,25 +17,17 @@ class ARRAY(UserDefinedType):
         Args:
             item_type: The data type of items in this array. For nested arrays,
                       pass another ARRAY type.
-
-        Raises:
-            ValueError: If nesting level exceeds the maximum allowed level (6).
         """
         super(UserDefinedType, self).__init__()
         if isinstance(item_type, type):
             item_type = item_type()
         self.item_type = item_type
-        self._validate_nesting_level()
-
-    def _validate_nesting_level(self):
-        """Validate that the nesting level does not exceed the maximum allowed level."""
-        level = 1
-        current_type = self.item_type
-        while isinstance(current_type, ARRAY):
-            level += 1
-            if level > self._max_nesting_level:
-                raise ValueError(f"Maximum nesting level of {self._max_nesting_level} exceeded")
-            current_type = current_type.item_type
+        if isinstance(item_type, ARRAY):
+            self.dim = item_type.dim + 1
+        else:
+            self.dim = 1
+        if self.dim > 6:
+            raise ValueError("Maximum nesting level of 6 exceeded")
 
     def get_col_spec(self, **kw):  # pylint: disable=unused-argument
         """Parse to array data type definition in text SQL."""
@@ -46,12 +37,33 @@ class ARRAY(UserDefinedType):
             base_type = str(self.item_type)
         return f"ARRAY({base_type})"
 
+    def _get_list_depth(self, value: Any) -> int:
+        if not isinstance(value, list):
+            return 0
+        max_depth = 0
+        for element in value:
+            current_depth = self._get_list_depth(element)
+            if current_depth > max_depth:
+                max_depth = current_depth
+        return 1 + max_depth
+
+    def _validate_dimension(self, value: list[Any]):
+        arr_depth = self._get_list_depth(value)
+        assert arr_depth == self.dim, "Array dimension mismatch, expected {}, got {}".format(self.dim, arr_depth)
+
     def bind_processor(self, dialect):
-        item_proc = self.item_type.dialect_impl(dialect).bind_processor(dialect)
+        item_type = self.item_type
+        while isinstance(item_type, ARRAY):
+            item_type = item_type.item_type
 
-        def process(value: Optional[Sequence[Any]]) -> Optional[str]:
+        item_proc = item_type.dialect_impl(dialect).bind_processor(dialect)
+
+        def process(value: Optional[Sequence[Any] | str]) -> Optional[str]:
             if value is None:
                 return None
+            if isinstance(value, str):
+                self._validate_dimension(json.loads(value))
+                return value
 
             def convert(val):
                 if isinstance(val, (list, tuple)):
@@ -61,12 +73,17 @@ class ARRAY(UserDefinedType):
                 return val
 
             processed = convert(value)
+            self._validate_dimension(processed)
             return json.dumps(processed)
 
         return process
 
     def result_processor(self, dialect, coltype):
-        item_proc = self.item_type.dialect_impl(dialect).result_processor(dialect, coltype)
+        item_type = self.item_type
+        while isinstance(item_type, ARRAY):
+            item_type = item_type.item_type
+
+        item_proc = item_type.dialect_impl(dialect).result_processor(dialect, coltype)
 
         def process(value: Optional[str]) -> Optional[List[Any]]:
             if value is None:
@@ -85,7 +102,11 @@ class ARRAY(UserDefinedType):
         return process
 
     def literal_processor(self, dialect):
-        item_proc = self.item_type.dialect_impl(dialect).literal_processor(dialect)
+        item_type = self.item_type
+        while isinstance(item_type, ARRAY):
+            item_type = item_type.item_type
+
+        item_proc = item_type.dialect_impl(dialect).literal_processor(dialect)
 
         def process(value: Sequence[Any]) -> str:
             def convert(val):
@@ -100,21 +121,6 @@ class ARRAY(UserDefinedType):
 
         return process
 
-    def __repr__(self):
-        """Return a string representation of the array type."""
-        current_type = self.item_type
-        nesting_level = 1
-        base_type = current_type
-
-        # Find the innermost type and count nesting level
-        while isinstance(current_type, ARRAY):
-            nesting_level += 1
-            current_type = current_type.item_type
-            if not isinstance(current_type, ARRAY):
-                base_type = current_type
-
-        return f"{nesting_level}D_Array({base_type})"
-
 
 def nested_array(dim: int) -> Type[ARRAY]:
     """Create a nested array type class with specified dimensions.
@@ -131,12 +137,22 @@ def nested_array(dim: int) -> Type[ARRAY]:
     if not 1 <= dim <= 6:
         raise ValueError("Dimension must be between 1 and 6")
 
-    class ArrayType(ARRAY):
+    class NestedArray(ARRAY):
+        cache_ok = True
+        _string = String()
+
         def __init__(self, item_type: Union[TypeEngine, type]):
+            super(UserDefinedType, self).__init__()
+            if isinstance(item_type, type):
+                item_type = item_type()
+
+            assert not isinstance(item_type, ARRAY), "The item_type of NestedArray should not be an ARRAY type"
+
             nested_type = item_type
-            for _ in range(dim - 1):
+            for _ in range(dim):
                 nested_type = ARRAY(nested_type)
-            super().__init__(nested_type)
 
-    ArrayType.__name__ = f"{dim}D_Array"
-    return ArrayType
+            self.item_type = nested_type.item_type
+            self.dim = dim
+
+    return NestedArray

{pyobvector-0.2.12.dist-info → pyobvector-0.2.13.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: pyobvector
-Version: 0.2.12
+Version: 0.2.13
 Summary: A python SDK for OceanBase Vector Store, based on SQLAlchemy, compatible with Milvus API.
 Author: shanhaikang.shk
 Author-email: shanhaikang.shk@oceanbase.com
@@ -36,7 +36,7 @@ poetry install
 - install with pip:
 
 ```shell
-pip install pyobvector==0.2.12
+pip install pyobvector==0.2.13
 ```
 
 ## Build Doc

{pyobvector-0.2.12.dist-info → pyobvector-0.2.13.dist-info}/RECORD

@@ -6,7 +6,7 @@ pyobvector/client/exceptions.py,sha256=CAsTHR9juYleRjYIa4bqk_lw14h8daBvChKoU0o19
 pyobvector/client/fts_index_param.py,sha256=hMCjA3Aecnt0uQQT6UQGTIIqdPk1M4gX4-zREDQygLs,1139
 pyobvector/client/index_param.py,sha256=3gXi66Ey1PO9x5_61CrH7DmPb496kviBQI5NT7nfbGc,6309
 pyobvector/client/milvus_like_client.py,sha256=CpPo6mkGE8iNFpKGBFof3h7E1VTzy1DAPGlFM9F_s8g,26373
-pyobvector/client/ob_vec_client.py,sha256=Yt2nG0w4268hg7DE0tqkGaytGsY-jqojX8hGTQjmsKg,29390
+pyobvector/client/ob_vec_client.py,sha256=QZAZxtSgsS2Z8SgEtsSSUcCroaML_XS-cq7aK-sSxic,29718
 pyobvector/client/ob_vec_json_table_client.py,sha256=m0Oq41dXEil9S1YCK2_RGbSMziatqAItn8Osk-9rzJI,39066
 pyobvector/client/partitions.py,sha256=Bxwr5yVNlXwZc7SXBC03NeqL9giy4Fe6S2qZdHD8xGw,15621
 pyobvector/client/schema_type.py,sha256=u1LJsr1o9lxv2b_6KYu77RciFa1R_Qk69k_WT30x6BU,1582
@@ -15,7 +15,7 @@ pyobvector/json_table/json_value_returning_func.py,sha256=NWSV2zhe2-1KhIprQaFqOH
 pyobvector/json_table/oceanbase_dialect.py,sha256=lxpbWBQdK18LWXLmGyk_-ODv6VfnwGLHbcpsQMElOUo,4480
 pyobvector/json_table/virtual_data_type.py,sha256=uQh6ZQ0UbwpVO9TFegGeu4E8bXW7rdLHAXFQJdiEjLs,3467
 pyobvector/schema/__init__.py,sha256=EU8NH8Q-L05sFBGKPV6yIBUeh5f3awTkArdBJ7d4CvQ,2271
-pyobvector/schema/array.py,sha256=cFAbayxsJoArPciawFImH835Wnrfm6KeahjBQB7uS44,4787
+pyobvector/schema/array.py,sha256=PoSBc3qCVdMJcLramZp95t69i15ES1J_bqnFANqQoRs,5255
 pyobvector/schema/dialect.py,sha256=mdRjn3roztCkk6RXbaB0Wn1uhT2BPS2y18MwL6wW-jo,1840
 pyobvector/schema/full_text_index.py,sha256=ohQX8uTPdRswEJONuN5A-bNv203d0N0b2BsJ7etx71g,2071
 pyobvector/schema/geo_srid_point.py,sha256=RwEoCgGTmXDc0le1B2E3mZudtqiFdMf2M0Va1ocmVSY,1210
@@ -30,7 +30,7 @@ pyobvector/schema/vector_index.py,sha256=aNtrEBUclc4s6QuqCZpu3Hj3OdjyhbWgtLiJzo6
 pyobvector/util/__init__.py,sha256=D9EgRDlcMSDhY3uI__vnCl45Or75dOXMWSval5P5fqs,251
 pyobvector/util/ob_version.py,sha256=ZIySam8q_MCiwctAiAHPB4GdAzGQiXEo1wVkc9IOTDU,1539
 pyobvector/util/vector.py,sha256=xyM-NuOyd78K7P3kinqyWvLIzEbf9c-4TKn_QVF7qgw,2265
-pyobvector-0.2.12.dist-info/LICENSE,sha256=WNHhf_5RCaeuKWyq_K39vmp9F28LxKsB4SpomwSZ2L0,11357
-pyobvector-0.2.12.dist-info/METADATA,sha256=b8nShwi2i0RfIRXdmaF-9zgy3LTt0mpujD3tgtOx9L8,6659
-pyobvector-0.2.12.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
-pyobvector-0.2.12.dist-info/RECORD,,
+pyobvector-0.2.13.dist-info/LICENSE,sha256=WNHhf_5RCaeuKWyq_K39vmp9F28LxKsB4SpomwSZ2L0,11357
+pyobvector-0.2.13.dist-info/METADATA,sha256=fxRXFWsunoyvBZ3wW5r7FMo8JB6WZbYBj9ijOtF8tnU,6659
+pyobvector-0.2.13.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+pyobvector-0.2.13.dist-info/RECORD,,