ingest_rag_data.py

import argparse
import json
import logging
from pathlib import Path


def build_parser() -> argparse.ArgumentParser:
    parser = argparse.ArgumentParser(
        description="Ingest medical documents into the configured RAG backend."
    )
    source = parser.add_mutually_exclusive_group(required=True)
    source.add_argument("--file", type=Path, help="Path to one document to ingest")
    source.add_argument("--dir", type=Path, help="Directory of documents to ingest")
    return parser


def run_ingestion(args: argparse.Namespace) -> dict:
    # Keep optional RAG dependencies out of the CLI help path.
    from agents.rag_agent import MedicalRAG
    from config import Config

    rag = MedicalRAG(Config())
    if args.file is not None:
        return rag.ingest_file(str(args.file))
    return rag.ingest_directory(str(args.dir))


def main() -> int:
    logging.basicConfig(
        level=logging.INFO,
        format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
    )
    args = build_parser().parse_args()
    result = run_ingestion(args)
    print("Ingestion result:", json.dumps(result, indent=2))
    return 0 if result.get("success") else 1


if __name__ == "__main__":
    raise SystemExit(main())