INTPYTHON-810 - Create langchain-mongodb-retrievers

.github/scripts/check_diff.py

@@ -4,6 +4,7 @@
 
 LIB_DIRS = [
     "libs/langchain-mongodb",
+    "libs/langchain-mongodb-retrievers",
     "libs/langgraph-checkpoint-mongodb",
     "libs/langgraph-store-mongodb",
 ]

.github/workflows/_test.yml

@@ -18,7 +18,7 @@ jobs:
       fail-fast: false
       matrix:
         python-version:
-          - "3.9"
+          - "3.10"
           - "3.13"
     name: "run test #${{ matrix.python-version }}"
     steps:

libs/langchain-mongodb-retrievers/.gitignore

@@ -0,0 +1 @@
+__pycache__

libs/langchain-mongodb-retrievers/CHANGELOG.md

@@ -0,0 +1,7 @@
+# Changelog
+
+---
+
+## Changes in version 1.0 (2025/XX/XX)
+
+- Initial release, supporting LangChain 1.0.

libs/langchain-mongodb-retrievers/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 LangChain, Inc.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

libs/langchain-mongodb-retrievers/README.md

@@ -0,0 +1,38 @@
+from libs.community.tests.unit_tests.chains.test_pebblo_retrieval import retriever
+
+# langchain-mongodb-retriever
+
+# Installation
+```
+pip install -U langchain-mongodb-retriever
+```
+
+# Usage
+- [Integrate Atlas Vector Search with LangChain](https://www.mongodb.com/docs/atlas/atlas-vector-search/ai-integrations/langchain/#get-started-with-the-langchain-integration) for a walkthrough on using your first LangChain implementation with MongoDB Atlas.
+
+## Using MongoDBAtlasVectorSearch
+```python
+import os
+from langchain_mongodb_retrievers import MongoDBAtlasVectorSearch
+from langchain_openai import OpenAIEmbeddings
+
+# Pull MongoDB Atlas URI from environment variables
+MONGODB_ATLAS_CONNECTION_STRING = os.environ["MONGODB_CONNECTION_STRING"]
+DB_NAME = "langchain_db"
+COLLECTION_NAME = "test"
+VECTOR_SEARCH_INDEX_NAME = "index_name"
+
+MODEL_NAME = "text-embedding-3-large"
+OPENAI_API_KEY =  os.environ["OPENAI_API_KEY"]
+
+
+vectorstore = MongoDBAtlasVectorSearch.from_connection_string(
+    connection_string=MONGODB_ATLAS_CONNECTION_STRING,
+    namespace=DB_NAME + "." + COLLECTION_NAME,
+    embedding=OpenAIEmbeddings(model=MODEL_NAME),
+    index_name=VECTOR_SEARCH_INDEX_NAME,
+)
+
+retrieved_docs = vectorstore.similarity_search(
+    "How do I deploy MongoDBAtlasVectorSearch in our production environment?")
+```

libs/langchain-mongodb-retrievers/justfile

@@ -0,0 +1,40 @@
+set shell := ["bash", "-c"]
+set dotenv-load
+set dotenv-filename := "../../.local_atlas_uri"
+
+# Default target executed when no arguments are given.
+[private]
+default:
+  @just --list
+
+install:
+	uv sync --frozen
+
+[group('test')]
+integration_tests *args="":
+	uv run pytest tests/integration_tests/ {{args}}
+
+[group('test')]
+unit_tests *args="":
+	uv run pytest tests/unit_tests {{args}}
+
+[group('test')]
+tests *args="":
+	uv run pytest {{args}}
+
+[group('test')]
+test_watch filename:
+	uv run ptw --snapshot-update --now . -- -vv {{filename}}
+
+[group('lint')]
+lint:
+	git ls-files -- '*.py' | xargs uv run pre-commit run ruff --files
+	git ls-files -- '*.py' | xargs uv run pre-commit run ruff-format --files
+
+[group('lint')]
+typing:
+	uv run mypy --install-types --non-interactive .
+
+[group('lint')]
+codespell:
+	git ls-files -- '*.py' | xargs uv run pre-commit run --hook-stage manual codespell --files

libs/langchain-mongodb/langchain_mongodb/retrievers/__init__.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/__init__.py

@@ -4,20 +4,20 @@
 to create MongoDB's core Vector Search Retriever.
 """
 
-from langchain_mongodb.retrievers.full_text_search import (
+from langchain_mongodb_retrievers.full_text_search import (
     MongoDBAtlasFullTextSearchRetriever,
 )
-from langchain_mongodb.retrievers.graphrag import MongoDBGraphRAGRetriever
-from langchain_mongodb.retrievers.hybrid_search import MongoDBAtlasHybridSearchRetriever
-from langchain_mongodb.retrievers.parent_document import (
+from langchain_mongodb_retrievers.hybrid_search import MongoDBAtlasHybridSearchRetriever
+from langchain_mongodb_retrievers.parent_document import (
     MongoDBAtlasParentDocumentRetriever,
 )
-from langchain_mongodb.retrievers.self_querying import MongoDBAtlasSelfQueryRetriever
+from langchain_mongodb_retrievers.self_querying import MongoDBAtlasSelfQueryRetriever
+from langchain_mongodb_retrievers.vectorstores import MongoDBAtlasVectorSearch
 
 __all__ = [
     "MongoDBAtlasHybridSearchRetriever",
     "MongoDBAtlasFullTextSearchRetriever",
     "MongoDBAtlasParentDocumentRetriever",
-    "MongoDBGraphRAGRetriever",
     "MongoDBAtlasSelfQueryRetriever",
+    "MongoDBAtlasVectorSearch",
 ]

libs/langchain-mongodb/langchain_mongodb/cache.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/cache.py

@@ -10,12 +10,12 @@
 from langchain_core.load.dump import dumps
 from langchain_core.load.load import loads
 from langchain_core.outputs import Generation
+from langchain_mongodb.utils import DRIVER_METADATA
 from pymongo import MongoClient
 from pymongo.collection import Collection
 from pymongo.database import Database
 
-from langchain_mongodb.utils import DRIVER_METADATA
-from langchain_mongodb.vectorstores import MongoDBAtlasVectorSearch
+from langchain_mongodb_retrievers.vectorstores import MongoDBAtlasVectorSearch
 
 logger = logging.getLogger(__file__)
 

libs/langchain-mongodb/langchain_mongodb/retrievers/full_text_search.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/full_text_search.py

@@ -4,11 +4,10 @@
 from langchain_core.callbacks.manager import CallbackManagerForRetrieverRun
 from langchain_core.documents import Document
 from langchain_core.retrievers import BaseRetriever
-from pydantic import Field
-from pymongo.collection import Collection
-
 from langchain_mongodb.pipelines import text_search_stage
 from langchain_mongodb.utils import _append_client_metadata, make_serializable
+from pydantic import Field
+from pymongo.collection import Collection
 
 
 class MongoDBAtlasFullTextSearchRetriever(BaseRetriever):

libs/langchain-mongodb/langchain_mongodb/retrievers/hybrid_search.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/hybrid_search.py

@@ -4,10 +4,6 @@
 from langchain_core.callbacks.manager import CallbackManagerForRetrieverRun
 from langchain_core.documents import Document
 from langchain_core.retrievers import BaseRetriever
-from pydantic import Field
-from pymongo.collection import Collection
-
-from langchain_mongodb import MongoDBAtlasVectorSearch
 from langchain_mongodb.pipelines import (
     combine_pipelines,
     final_hybrid_stage,
@@ -16,6 +12,10 @@
     vector_search_stage,
 )
 from langchain_mongodb.utils import make_serializable
+from pydantic import Field
+from pymongo.collection import Collection
+
+from langchain_mongodb_retrievers.vectorstores import MongoDBAtlasVectorSearch
 
 
 class MongoDBAtlasHybridSearchRetriever(BaseRetriever):

libs/langchain-mongodb/langchain_mongodb/retrievers/parent_document.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/parent_document.py

@@ -3,21 +3,23 @@
 from typing import Any, List, Optional
 
 import pymongo
-from langchain.retrievers.parent_document_retriever import ParentDocumentRetriever
+from langchain_classic.retrievers.parent_document_retriever import (
+    ParentDocumentRetriever,
+)
 from langchain_core.callbacks import (
     AsyncCallbackManagerForRetrieverRun,
     CallbackManagerForRetrieverRun,
 )
 from langchain_core.documents import Document
 from langchain_core.embeddings import Embeddings
 from langchain_core.runnables import run_in_executor
-from langchain_text_splitters import TextSplitter
-from pymongo import MongoClient
-
-from langchain_mongodb import MongoDBAtlasVectorSearch
 from langchain_mongodb.docstores import MongoDBDocStore
 from langchain_mongodb.pipelines import vector_search_stage
 from langchain_mongodb.utils import DRIVER_METADATA, make_serializable
+from langchain_text_splitters import TextSplitter
+from pymongo import MongoClient
+
+from langchain_mongodb_retrievers.vectorstores import MongoDBAtlasVectorSearch
 
 
 class MongoDBAtlasParentDocumentRetriever(ParentDocumentRetriever):
@@ -39,8 +41,8 @@ class MongoDBAtlasParentDocumentRetriever(ParentDocumentRetriever):
     by the same MongoDB Collection.
 
     For more details, see superclasses
-        :class:`~langchain.retrievers.parent_document_retriever.ParentDocumentRetriever`
-        and :class:`~langchain.retrievers.MultiVectorRetriever`.
+        :class:`~langchain_classic.retrievers.parent_document_retriever.ParentDocumentRetriever`
+        and :class:`~langchain_classic.retrievers.MultiVectorRetriever`.
 
     Examples:
         >>> from langchain_mongodb.retrievers.parent_document import (
@@ -148,8 +150,8 @@ def from_connection_string(
         """Construct Retriever using one Collection for VectorStore and one for DocStore
 
         See parent classes
-        :class:`~langchain.retrievers.parent_document_retriever.ParentDocumentRetriever`
-        and :class:`~langchain.retrievers.MultiVectorRetriever` for further details.
+        :class:`~langchain_classic.retrievers.parent_document_retriever.ParentDocumentRetriever`
+        and :class:`~langchain_classic.retrievers.MultiVectorRetriever` for further details.
 
         Args:
             connection_string: A valid MongoDB Atlas connection URI.

libs/langchain-mongodb/langchain_mongodb/retrievers/self_querying.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/self_querying.py

@@ -1,7 +1,7 @@
 from typing import Any, Dict, Sequence, Tuple, Union
 
-from langchain.chains.query_constructor.schema import AttributeInfo
-from langchain.retrievers.self_query.base import SelfQueryRetriever
+from langchain_classic.chains.query_constructor.schema import AttributeInfo
+from langchain_classic.retrievers.self_query.base import SelfQueryRetriever
 from langchain_core.language_models import BaseLanguageModel
 from langchain_core.runnables import Runnable
 from langchain_core.structured_query import (
@@ -15,7 +15,7 @@
 from langchain_core.vectorstores import VectorStore
 from pydantic import Field
 
-from langchain_mongodb import MongoDBAtlasVectorSearch
+from langchain_mongodb_retrievers.vectorstores import MongoDBAtlasVectorSearch
 
 
 class MongoDBStructuredQueryTranslator(Visitor):

libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/utils.py

@@ -0,0 +1,21 @@
+"""Various Utility Functions"""
+
+from __future__ import annotations
+
+import logging
+from importlib.metadata import version
+
+from pymongo import MongoClient
+from pymongo.driver_info import DriverInfo
+
+logger = logging.getLogger(__name__)
+
+DRIVER_METADATA = DriverInfo(
+    name="langchain-mongodb-retrievers", version=version("langchain-mongodb-retrievers")
+)
+
+
+def _append_client_metadata(client: MongoClient) -> None:
+    # append_metadata was added in PyMongo 4.14.0, but is a valid database name on earlier versions
+    if callable(client.append_metadata):
+        client.append_metadata(DRIVER_METADATA)

libs/langchain-mongodb/langchain_mongodb/vectorstores.py → libs/langchain-mongodb-retrievers/langchain_mongodb_retrievers/vectorstores.py

@@ -22,10 +22,6 @@
 from langchain_core.embeddings import Embeddings
 from langchain_core.runnables.config import run_in_executor
 from langchain_core.vectorstores import VectorStore
-from pymongo import MongoClient, ReplaceOne
-from pymongo.collection import Collection
-from pymongo.errors import CollectionInvalid
-
 from langchain_mongodb.index import (
     create_vector_search_index,
     update_vector_search_index,
@@ -39,6 +35,9 @@
     oid_to_str,
     str_to_oid,
 )
+from pymongo import MongoClient, ReplaceOne
+from pymongo.collection import Collection
+from pymongo.errors import CollectionInvalid
 
 VST = TypeVar("VST", bound=VectorStore)
 
@@ -356,7 +355,7 @@ def add_texts(
             metadatas_batch = []
             size = 0
             i = 0
-            for j, (text, metadata) in enumerate(zip(texts, _metadatas)):
+            for j, (text, metadata) in enumerate(zip(texts, _metadatas, strict=False)):
                 size += len(text) + len(metadata)
                 texts_batch.append(text)
                 metadatas_batch.append(metadata)
@@ -442,7 +441,9 @@ def bulk_embed_and_insert_texts(
                 self._embedding_key: embedding,
                 **m,
             }
-            for i, t, m, embedding in zip(ids, texts, metadatas, embeddings)
+            for i, t, m, embedding in zip(
+                ids, texts, metadatas, embeddings, strict=False
+            )
         ]
         operations = [ReplaceOne({"_id": doc["_id"]}, doc, upsert=True) for doc in docs]
         # insert the documents in MongoDB Atlas
@@ -478,7 +479,8 @@ def add_documents(
         start = 0
         for end in range(batch_size, n_docs + batch_size, batch_size):
             texts, metadatas = zip(
-                *[(doc.page_content, doc.metadata) for doc in documents[start:end]]
+                *[(doc.page_content, doc.metadata) for doc in documents[start:end]],
+                strict=False,
             )
             result_ids.extend(
                 self.bulk_embed_and_insert_texts(