scripts/build_index.py

import argparse
import sys
import os

sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))

from dotenv import load_dotenv
load_dotenv()

from retrieval.ingest import (
    ingest_all_dataset_files,
    clear_and_rebuild_index,
    ingest_folder
)
from retrieval.faiss_store import index_size

def main():
    parser = argparse.ArgumentParser(description="Build AccessoryIQ FAISS index")
    parser.add_argument("--rebuild", action="store_true",
                        help="Clear existing index and rebuild from scratch")
    parser.add_argument("--folder",  type=str, default=None,
                        help="Ingest only a specific subfolder")
    parser.add_argument("--sample",  action="store_true",
                        help="Also ingest sample data from ingest.py")
    args = parser.parse_args()

    print("\n" + "─"*50)
    print(" AccessoryIQ v2 — Index Builder")
    print("─"*50)
    print(f" Current index size: {index_size()} vectors")
    print("─"*50 + "\n")

    if args.rebuild:
        clear_and_rebuild_index()
    elif args.folder:
        path = f"data/manuals/{args.folder}"
        if not os.path.exists(path):
            print(f"[ERROR] Folder not found: {path}")
            sys.exit(1)
        ingest_folder(path)
    else:
        if args.sample:
            from retrieval.ingest import ingest_sample_data
            ingest_sample_data()
        ingest_all_dataset_files()

    final_size = index_size()
    print(f"\n{'─'*50}")
    print(f" ✓ Index build complete: {final_size} vectors")
    print(f"─"*50 + "\n")

if __name__ == "__main__":
    main()