"""Build the vector store from data/dataset/chunks.jsonl.

    python scripts/build_vectorstore.py                      # tfidf (offline default)
    python scripts/build_vectorstore.py --backend sbert      # dense embeddings
    python scripts/build_vectorstore.py --backend openai     # OpenAI-compatible API
"""

import argparse
import sys
from pathlib import Path

sys.path.insert(0, str(Path(__file__).resolve().parent))
from vectorstore import build_index  # noqa: E402

ROOT = Path(__file__).resolve().parent.parent

if __name__ == "__main__":
    ap = argparse.ArgumentParser()
    ap.add_argument("--chunks", default=str(ROOT / "data" / "dataset" / "chunks.jsonl"))
    ap.add_argument("--out", default=str(ROOT / "vectorstore"))
    ap.add_argument("--backend", default="tfidf", choices=["tfidf", "sbert", "openai"])
    ap.add_argument("--provider", default=None, help="cloud embedding provider preset, e.g. siliconflow / zhipu / qwen")
    ap.add_argument("--model", default=None)
    args = ap.parse_args()
    build_index(args.chunks, args.out, backend=args.backend, model=args.model, provider=args.provider)