# app.py

import gradio as gr
from pipeline import chatbot_response, generate_coherent_response, ChatHistory

# Create a new instance of chat history
chat_history = ChatHistory()

# Step 6: Define Gradio Interface for Chatbot
def gradio_chatbot(user_query):
    # Step 7: Retrieve relevant data for the query
    retrieved_data = chatbot_response(user_query)

    # Step 8: Check and verify the query for health/wellness content
    is_valid = True
    if is_valid:
        # Generate a coherent response using Groq's DeepSeek-R1 LLM
        coherent_response = generate_coherent_response(user_query, retrieved_data, chat_history.get_history())
    else:
        coherent_response = "The query does not seem to match health and wellness topics."

    # Add the user message and assistant response to the chat history
    chat_history.add_message("user", user_query)
    chat_history.add_message("assistant", coherent_response)

    # Return the response and the chat history
    return coherent_response, chat_history.get_history()

# Create the Gradio interface
iface = gr.Interface(fn=gradio_chatbot,
                     inputs=gr.Textbox(label="Ask a question"),
                     outputs=[gr.Textbox(label="Chatbot Response"), gr.Textbox(label="Chat History")],
                     title="Wellness Chatbot")

# Launch the Gradio interface
iface.launch()