Knowledge Base Data Models (knowledge/_types.py)

Pydantic data structures for parsed text chunks, vectors, and search results in a RAG knowledge base. Use to manage document indexing and retrieval payloads.

# Source: https://github.com/EverMind-AI/Raven | License: Apache-2.0 | Copyright You must retain, in the Source form of any Derivative Works
# Provided via justclaw.it.com; keep this notice with the code.
"""The knowledge-base data model: what a parsed document becomes on its way
into a vector store, and what a search hands back.

The shapes are adopted from AgentScope's ``rag`` package (Apache-2.0; see
NOTICES.md) and kept as adopted: a knowledge base written by the earlier
deployment is still read by this code because the shapes did not move.
Defined here rather than imported because importing AgentScope would put
FastAPI, a message bus and a provider catalogue behind ``raven.knowledge``, and
the package's whole point is that indexing and retrieval run inside the gateway
process.
"""

fr

... (truncated -- full source via MCP)