main.py

import os
import sys
from pathlib import Path

# Add src directory to path
sys.path.append(os.path.join(os.path.dirname(__file__), 'src'))

from qdrant_setup import QdrantSetup
from document_ingestor import DocumentIngestor
from load_huggingface_dataset import HuggingFaceDatasetLoader
from embedding_generator import get_embedding_function


def setup_and_run():
    """
    Main function to setup and run the Hindi RAG system
    """
    print("Setting up Hindi RAG System...")

    # Step 1: Initialize Qdrant
    print("\n1. Setting up Qdrant...")
    qdrant_setup = QdrantSetup()
    qdrant_setup.create_collection()

    # Step 2: Load and ingest documents from HuggingFace datasets
    print("\n2. Loading datasets from HuggingFace...")
    qdrant_client = qdrant_setup.get_client()
    collection_name = qdrant_setup.get_collection_name()
    embedding_func = get_embedding_function()

    hf_loader = HuggingFaceDatasetLoader(qdrant_client, collection_name)
    num_hf_docs = hf_loader.load_and_ingest_all_configured(embedding_func)
    if num_hf_docs > 0:
        print(f"   ✓ Loaded {num_hf_docs} new documents from HuggingFace datasets")
    else:
        hf_datasets = os.getenv("HF_DATASETS", "")
        if hf_datasets:
            print(f"   ⊘ HuggingFace datasets already ingested (skipped)")
        else:
            print("   No HuggingFace datasets configured (set HF_DATASETS env variable)")

    # Step 3: Load and ingest local documents
    print("\n3. Loading and ingesting local documents...")
    ingestor = DocumentIngestor(qdrant_client, collection_name)

    # Use the data directory
    data_dir = "./data"
    if os.path.exists(data_dir):
        num_docs = ingestor.load_and_ingest(data_dir, embedding_func)
        print(f"   Loaded {num_docs} local documents into Qdrant")
    else:
        print(f"   Warning: Data directory '{data_dir}' does not exist")

    # Step 4: Initialize RAG system with LLM provider from environment
    print("\n4. Initializing RAG system...")
    llm_provider = os.getenv("LLM_PROVIDER", "huggingface")
    print(f"   Using LLM provider: {llm_provider}")
    print("   (RAG system will be initialized by frontend on first request)")

    print("\n5. Hindi RAG System setup complete!")
    print("   Starting Streamlit frontend...")

if __name__ == "__main__":
    setup_and_run()