Added fetch API for vector index

Author: aosingh

doc/source/user_guide/installation.rst

@@ -11,7 +11,12 @@ Installation requirements
 
 To use ``select_ai`` you need:
 
-- Python 3.9, 3.10, 3.11, 3.12 or 3.13
+- Python 3.9, 3.10, 3.11, 3.12, 3.13 or 3.14
+
+.. warning::
+For async APIs, use Python 3.11 or higher. Python 3.11 introduced
+stabilized the async event loop management and introduced better-structured
+APIs
 
 - ``python-oracledb`` - This package is automatically installed as a dependency requirement
 
@@ -28,7 +33,7 @@ To use ``select_ai`` you need:
 `pip <https://pip.pypa.io/en/latest/installation/>`__.
 
 1. Install `Python 3 <https://www.python.org/downloads>`__ if it is not already
-   available. Use any version from Python 3.9 through 3.13.
+   available. Use any version from Python 3.9 through 3.14.
 
 2. Install ``select_ai``:
 

doc/source/user_guide/vector_index.rst

@@ -86,10 +86,9 @@ Get vector index attributes
 +++++++++++++++++++++++++++
 
 You can fetch the vector index attributes and associated AI profile using
-``vector_index.get_attributes()`` and ``vector_index.get_profile()`` methods
-respectively.
+the class method ``VectorIndex.fetch(index_name)``
 
-.. literalinclude:: ../../../samples/vector_index_get_attributes.py
+.. literalinclude:: ../../../samples/vector_index_fetch.py
    :language: python
    :lines: 14-
 
@@ -200,10 +199,9 @@ Async get vector index attributes
 +++++++++++++++++++++++++++++++++
 
 You can fetch the vector index attributes and associated AI profile using
-``async_vector_index.get_attributes()`` and ``async_vector_index.get_profile()``
-methods respectively.
+the class method ``AsyncVectorIndex.fetch(index_name)``
 
-.. literalinclude:: ../../../samples/async/vector_index_get_attributes.py
+.. literalinclude:: ../../../samples/async/vector_index_fetch.py
    :language: python
    :lines: 14-
 

samples/async/vector_index_get_attributes.py renamed to samples/async/vector_index_fetch.py

@@ -6,7 +6,7 @@
 # -----------------------------------------------------------------------------
 
 # -----------------------------------------------------------------------------
-# async/vector_index_get_attributes.py
+# async/vector_index_fetch.py
 #
 # Get vector index attributes
 # -----------------------------------------------------------------------------
@@ -23,14 +23,11 @@
 
 async def main():
     await select_ai.async_connect(user=user, password=password, dsn=dsn)
-
-    async_vector_index = select_ai.AsyncVectorIndex(
-        index_name="test_vector_index",
+    async_vector_index = await select_ai.AsyncVectorIndex.fetch(
+        index_name="test_vector_index"
     )
-    attributes = await async_vector_index.get_attributes()
-    print(attributes)
-    async_profile = await async_vector_index.get_profile()
-    print(async_profile)
+    print(async_vector_index.attributes)
+    print(async_vector_index.profile)
 
 
 asyncio.run(main())

samples/async/vector_index_update_attributes.py

@@ -29,16 +29,15 @@ async def main():
 
     # Use vector_index.set_attributes to update a multiple attributes
     updated_attributes = select_ai.OracleVectorIndexAttributes(
-        refresh_rate=1450,
+        refresh_rate=1450
     )
     await async_vector_index.set_attributes(attributes=updated_attributes)
 
     # Use vector_index.set_attribute to update a single attribute
     await async_vector_index.set_attribute(
         attribute_name="similarity_threshold", attribute_value=0.5
     )
-    attributes = await async_vector_index.get_attributes()
-    print(attributes)
+    print(async_vector_index.attributes)
 
 
 asyncio.run(main())

samples/vector_index_get_attributes.py renamed to samples/vector_index_fetch.py

@@ -6,7 +6,7 @@
 # -----------------------------------------------------------------------------
 
 # -----------------------------------------------------------------------------
-# vector_index_get_attributes.py
+# vector_index_fetch.py
 #
 # Gets attributes for a vector index
 # -----------------------------------------------------------------------------
@@ -21,9 +21,7 @@
 
 select_ai.connect(user=user, password=password, dsn=dsn)
 
-vector_index = select_ai.VectorIndex(
-    index_name="test_vector_index",
-)
-
-print(vector_index.get_attributes())
-print(vector_index.get_profile())
+vector_index = select_ai.VectorIndex.fetch(index_name="test_vector_index")
+print(vector_index.attributes)
+print(vector_index.profile)
+print(select_ai.VectorIndex(index_name=None).get_attributes())

samples/vector_index_update_attributes.py

@@ -24,13 +24,11 @@
 )
 
 # Use vector_index.set_attributes to update a multiple attributes
-updated_attributes = select_ai.OracleVectorIndexAttributes(
-    refresh_rate=1450,
-)
+updated_attributes = select_ai.OracleVectorIndexAttributes(refresh_rate=1450)
 vector_index.set_attributes(attributes=updated_attributes)
 
 # Use vector_index.set_attribute to update a single attribute
 vector_index.set_attribute(
     attribute_name="similarity_threshold", attribute_value=0.5
 )
-print(vector_index.get_attributes())
+print(vector_index.attributes)

src/select_ai/profile.py

@@ -8,8 +8,7 @@
 import json
 from contextlib import contextmanager
 from dataclasses import replace as dataclass_replace
-from pprint import pformat
-from typing import Generator, Iterator, Mapping, Optional, Tuple, Union
+from typing import Generator, Mapping, Optional, Tuple, Union
 
 import oracledb
 import pandas

src/select_ai/sql.py

@@ -88,6 +88,12 @@
 WHERE REGEXP_LIKE(v.index_name, :index_name_pattern, 'i')
 """
 
+GET_USER_VECTOR_INDEX = """
+select index_name, description
+from USER_CLOUD_VECTOR_INDEXES v
+where index_name = :index_name
+"""
+
 GET_USER_VECTOR_INDEX_ATTRIBUTES = """
 SELECT attribute_name, attribute_value
 FROM USER_CLOUD_VECTOR_INDEX_ATTRIBUTES

src/select_ai/vector_index.py

@@ -20,20 +20,11 @@
 from select_ai.errors import ProfileNotFoundError, VectorIndexNotFoundError
 from select_ai.profile import Profile
 from select_ai.sql import (
+    GET_USER_VECTOR_INDEX,
     GET_USER_VECTOR_INDEX_ATTRIBUTES,
     LIST_USER_VECTOR_INDEXES,
 )
 
-UNMODIFIABLE_VECTOR_INDEX_ATTRIBUTES = (
-    "location",
-    "chunk_size",
-    "chunk_overlap",
-    "pipeline_name",
-    "vector_dimension",
-    "vector_table_name",
-    "vector_distance_metric",
-)
-
 
 class VectorDBProvider(StrEnum):
     ORACLE = "oracle"
@@ -161,7 +152,7 @@ def _get_attributes(index_name: str) -> VectorIndexAttributes:
         :return: select_ai.VectorIndexAttributes
         :raises: VectorIndexNotFoundError
         """
-        if index_name is None:
+        if not index_name:
             raise AttributeError("'index_name' is required")
         with cursor() as cr:
             cr.execute(
@@ -181,6 +172,27 @@ def _get_attributes(index_name: str) -> VectorIndexAttributes:
             else:
                 raise VectorIndexNotFoundError(index_name=index_name)
 
+    @staticmethod
+    def _get_description(index_name) -> Union[str, None]:
+        """Get description of the Vector Index from USER_CLOUD_VECTOR_INDEXES
+
+        :param str index_name: The name of the vector index
+        :return: Union[str, None] profile description
+        :raises: ProfileNotFoundError
+        """
+        if not index_name:
+            raise AttributeError("'index_name' is required")
+        with cursor() as cr:
+            cr.execute(GET_USER_VECTOR_INDEX, index_name=index_name.upper())
+            index = cr.fetchone()
+            if index:
+                if index[1] is not None:
+                    return index[1].read()
+                else:
+                    return None
+            else:
+                raise VectorIndexNotFoundError(index_name=index_name)
+
     def create(self, replace: Optional[bool] = False):
         """Create a vector index in the database and populates the index
          with data from an object store bucket using an async scheduler job
@@ -292,6 +304,28 @@ def disable(self):
                 else:
                     raise
 
+    @classmethod
+    def fetch(cls, index_name: str) -> "VectorIndex":
+        """
+        Fetches vector index attributes from the
+        database and builds a proxy object for the
+        passed index_name
+
+        :param str index_name: The name of the vector index
+        """
+        description = cls._get_description(index_name)
+        attributes = cls._get_attributes(index_name)
+        try:
+            profile = Profile(profile_name=attributes.profile_name)
+        except ProfileNotFoundError:
+            profile = None
+        return cls(
+            description=description,
+            attributes=attributes,
+            profile=profile,
+            index_name=index_name,
+        )
+
     def set_attribute(
         self,
         attribute_name: str,
@@ -377,21 +411,7 @@ def list(cls, index_name_pattern: str = ".*") -> Iterator["VectorIndex"]:
             )
             for row in cr.fetchall():
                 index_name = row[0]
-                if row[1]:
-                    description = row[1].read()  # Oracle.LOB
-                else:
-                    description = None
-                attributes = cls._get_attributes(index_name=index_name)
-                try:
-                    profile = Profile(profile_name=attributes.profile_name)
-                except ProfileNotFoundError:
-                    profile = None
-                yield cls(
-                    index_name=index_name,
-                    description=description,
-                    attributes=attributes,
-                    profile=profile,
-                )
+                yield cls.fetch(index_name=index_name)
 
 
 class AsyncVectorIndex(_BaseVectorIndex):
@@ -412,6 +432,8 @@ async def _get_attributes(index_name: str) -> VectorIndexAttributes:
         :return: select_ai.VectorIndexAttributes
         :raises: VectorIndexNotFoundError
         """
+        if not index_name:
+            raise AttributeError("'index_name' is required")
         async with async_cursor() as cr:
             await cr.execute(
                 GET_USER_VECTOR_INDEX_ATTRIBUTES, index_name=index_name.upper()
@@ -430,6 +452,29 @@ async def _get_attributes(index_name: str) -> VectorIndexAttributes:
             else:
                 raise VectorIndexNotFoundError(index_name=index_name)
 
+    @staticmethod
+    async def _get_description(index_name) -> Union[str, None]:
+        """Get description of the Vector Index from USER_CLOUD_VECTOR_INDEXES
+
+        :param str index_name: The name of the vector index
+        :return: Union[str, None] profile description
+        :raises: ProfileNotFoundError
+        """
+        if not index_name:
+            raise AttributeError("'index_name' is required")
+        async with async_cursor() as cr:
+            await cr.execute(
+                GET_USER_VECTOR_INDEX, index_name=index_name.upper()
+            )
+            index = await cr.fetchone()
+            if index:
+                if index[1] is not None:
+                    return await index[1].read()
+                else:
+                    return None
+            else:
+                raise VectorIndexNotFoundError(index_name=index_name)
+
     async def create(self, replace: Optional[bool] = False) -> None:
         """Create a vector index in the database and populates it with data
         from an object store bucket using an async scheduler job
@@ -539,6 +584,28 @@ async def disable(self) -> None:
                 else:
                     raise
 
+    @classmethod
+    async def fetch(cls, index_name: str) -> "AsyncVectorIndex":
+        """
+        Fetches vector index attributes from the
+        database and builds a proxy object for the
+        passed index_name
+
+        :param str index_name: The name of the vector index
+        """
+        description = await cls._get_description(index_name)
+        attributes = await cls._get_attributes(index_name)
+        try:
+            profile = await AsyncProfile(profile_name=attributes.profile_name)
+        except ProfileNotFoundError:
+            profile = None
+        return cls(
+            description=description,
+            attributes=attributes,
+            profile=profile,
+            index_name=index_name,
+        )
+
     async def set_attribute(
         self, attribute_name: str, attribute_value: Union[str, int, float]
     ) -> None:
@@ -605,7 +672,7 @@ async def get_profile(self) -> AsyncProfile:
     @classmethod
     async def list(
         cls, index_name_pattern: str = ".*"
-    ) -> AsyncGenerator[VectorIndex, None]:
+    ) -> AsyncGenerator["AsyncVectorIndex", None]:
         """List Vector Indexes.
 
         :param str index_name_pattern: Regular expressions can be used
@@ -623,20 +690,5 @@ async def list(
             rows = await cr.fetchall()
             for row in rows:
                 index_name = row[0]
-                if row[1]:
-                    description = await row[1].read()  # AsyncLOB
-                else:
-                    description = None
-                attributes = await cls._get_attributes(index_name=index_name)
-                try:
-                    profile = await AsyncProfile(
-                        profile_name=attributes.profile_name,
-                    )
-                except ProfileNotFoundError:
-                    profile = None
-                yield VectorIndex(
-                    index_name=index_name,
-                    description=description,
-                    attributes=attributes,
-                    profile=profile,
-                )
+                index = await cls.fetch(index_name=index_name)
+                yield index

tests/agents/test_3200_agents.py

@@ -34,8 +34,8 @@ def python_gen_ai_profile(profile_attributes):
 def agent_attributes():
     agent_attributes = AgentAttributes(
         profile_name=PYSAI_3200_PROFILE_NAME,
-        role="You are an AI Movie Analyst. "
-        "Your can help answer a variety of questions related to movies. ",
+        role="You are an AI Movie Analyst."
+        "You can help answer a variety of questions related to movies.",
         enable_human_tool=False,
     )
     return agent_attributes
@@ -69,3 +69,10 @@ def test_3201(agent_name_pattern):
     agent_descriptions = set(agent.description for agent in agents)
     assert PYSAI_3200_AGENT_NAME in agent_names
     assert PYSAI_3200_AGENT_DESCRIPTION in agent_descriptions
+
+
+def test_3203(agent_attributes):
+    agent = Agent.fetch(agent_name=PYSAI_3200_AGENT_NAME)
+    assert agent.agent_name == PYSAI_3200_AGENT_NAME
+    assert agent.attributes == agent_attributes
+    assert agent.description == PYSAI_3200_AGENT_DESCRIPTION