llama-index-vector-stores-cairn

LlamaIndex vector store for Cairn, the hybrid search database where a deletion is final.

pip install llama-index-vector-stores-cairn

from cairn_db import Client
from llama_index.core import StorageContext, VectorStoreIndex
from llama_index.vector_stores.cairn import CairnVectorStore, collection_schema

admin = Client("http://localhost:7200", api_key=ADMIN_KEY)
admin.create_collection("docs", collection_schema(1536, filterable={"source": "Enum"}))

store = CairnVectorStore(Client("http://localhost:7200", api_key=KEY).collection("docs"))
index = VectorStoreIndex.from_documents(documents, storage_context=StorageContext.from_defaults(vector_store=store))
index.as_retriever(similarity_top_k=4).retrieve("what changed in 2024?")
index.delete_ref_doc("handbook.pdf")   # the document and all its nodes, everywhere

• Storage. Each node is stored under its node id, with its source document in parent, so delete(ref_doc_id) is a single deletion by filter. The node's own serialization is kept, so nodes come back whole.
• Filters. Metadata filters on filterable fields: EQ, NE, GT, GTE, LT, LTE, IN, NIN, ANY, ALL, CONTAINS, IS_EMPTY, combined with AND, OR or NOT, and nested. doc_ids restricts the query to those source documents.
• Query modes. DEFAULT (vector), HYBRID (vector + BM25, fused) and TEXT_SEARCH (BM25).
• Not supported. TEXT_MATCH filters, and queries by node_ids (use get_nodes).
• Deletions are final. The client carries its consistency token, so later queries never return deleted nodes, on any node.