Load MongoDB Data for LlamaIndex

This integration provides a SimpleMongoReader for LlamaIndex, enabling the loading of data directly from a MongoDB database. It concatenates specified fields from each MongoDB document into a single document, making the data usable by LlamaIndex.

Use this reader when you need to ingest data stored in MongoDB into your LlamaIndex pipelines. Do not use this reader if your data is not stored in MongoDB or if you do not intend to use it with the LlamaIndex framework.

The reader takes connection details for your MongoDB instance (host, port, or URI), database name, collection name, and optionally, field names to concatenate, a separator, a query dictionary for filtering, a maximum number of documents to load, and metadata field names. The output is a list of documents, where each document contains the concatenated content from the specified fields of a MongoDB record.

This loader is designed to be used as a way to load data into LlamaIndex and/or subsequently used as a Tool in a LangChain Agent.

This integration is for developers working with LlamaIndex and MongoDB.

pip install llama-index-readers-mongodb

from llama_index.readers.mongodb import SimpleMongoReader

### Initialize SimpleMongoReader
reader = SimpleMongoReader(
    host="<Mongo Host>",  # Mongo host address
    port=27017,  # Mongo port (default: 27017)
    uri="<Mongo Connection String>",  # Provide the URI if not using host and port
)

### Lazy load data from MongoDB
documents = reader.lazy_load_data(
    db_name="<Database Name>",  # Name of the database
    collection_name="<Collection Name>",  # Name of the collection
    field_names=[
        "text"
    ],  # Names of the fields to concatenate (default: ["text"])
    separator="",  # Separator between fields (default: "")
    query_dict=None,  # Query to filter documents (default: None)
    max_docs=0,  # Maximum number of documents to load (default: 0)
    metadata_names=None,  # Names of the fields to add to metadata attribute (default: None)
)