Reindexer Vector Store for LangChain

This package provides a vector store integration for Reindexer database with the LangChain framework.

Installation

pip install langchain-reindexer

Usage

Now you can use the vector store in your LangChain application:

from langchain_reindexer import ReindexerVectorStore
from langchain_openai import OpenAIEmbeddings

# Initialize the vector store
vector_store = ReindexerVectorStore(
    embedding=OpenAIEmbeddings(),
    rx_connector_config={"dsn": "builtin:///tmp/my_db"},
    rx_namespace="my_namespace",
)

# Add documents
from langchain_core.documents import Document

documents = [
    Document(page_content="foo", metadata={"baz": "bar"}),
    Document(page_content="thud", metadata={"bar": "baz"}),
]

ids = vector_store.add_documents(documents=documents)

# Search
results = vector_store.similarity_search(query="thud", k=1)

More examples here

Features

• Add and delete documents
• Similarity search with and without scores
• Metadata filtering
• Maximal Marginal Relevance (MMR) search
• Async support
• Save and load vector store configuration