app.py

import gradio as gr
from gtts import gTTS
import os
import speech_recognition as sr

# Menu for the restaurant
menu = {
    "Starters": ["Soup", "Spring Rolls"],
    "Main Course": ["Paneer Butter Masala", "Chicken Curry", "Veg Biryani"],
    "Breads": ["Roti", "Naan", "Paratha"],
    "Desserts": ["Gulab Jamun", "Ice Cream"],
    "Drinks": ["Mango Lassi", "Soda", "Water"]
}

# Conversation history and context
conversation_history = []
context = {"last_action": None, "order": []}

# Text-to-Speech Function
def speak_and_save(text, filename="response.mp3"):
    tts = gTTS(text=text, lang='en')
    tts.save(filename)
    return filename

# Speech Recognition Function
def recognize_speech(audio_file_path):
    recognizer = sr.Recognizer()
    try:
        with sr.AudioFile(audio_file_path) as source:
            audio_data = recognizer.record(source)
            text = recognizer.recognize_google(audio_data)
            return text
    except sr.UnknownValueError:
        return "I didn't catch that. Could you repeat?"
    except sr.RequestError:
        return "There was an issue processing your speech. Try again later."

# Voice Assistant Logic
def process_order(audio_file_path):
    user_input = recognize_speech(audio_file_path)
    conversation_history.append(f"Customer: {user_input}")
    
    if user_input.startswith("Sorry") or user_input.startswith("I didn't"):
        response = user_input
        conversation_history.append(f"AI: {response}")
        return response, speak_and_save(response)

    global context
    response = ""

    # Greet the customer
    if context["last_action"] is None:
        response = "Welcome to our restaurant. How can I assist you today?"
        context["last_action"] = "greet"
    
    # Show menu
    elif "menu" in user_input.lower():
        response = "Here is our menu:\n"
        for category, items in menu.items():
            response += f"{category}: {', '.join(items)}\n"
        response += "What would you like to order?"
        context["last_action"] = "show_menu"
    
    # Place an order
    elif "order" in user_input.lower() or any(item.lower() in user_input.lower() for item in sum(menu.values(), [])):
        for category, items in menu.items():
            for item in items:
                if item.lower() in user_input.lower():
                    context["order"].append(item)
                    break
        response = f"I've added {', '.join(context['order'])} to your order. Would you like anything else?"
        context["last_action"] = "place_order"
    
    # Finalize the order
    elif "no" in user_input.lower() or "that's it" in user_input.lower():
        response = f"Your final order is: {', '.join(context['order'])}. Thank you for your order. Your food will arrive shortly. Enjoy your meal!"
        conversation_history.append(f"Order Summary: {context['order']}")
        context["last_action"] = "final_order"
        context["order"] = []  # Reset the order
    
    # Mid-meal check
    elif "everything alright" in user_input.lower() or "anything else" in user_input.lower():
        response = "I hope you're enjoying your meal. Can I bring you anything else?"
        context["last_action"] = "check_meal"

    # Default response
    else:
        response = "I'm not sure what you meant. Could you clarify?"
        context["last_action"] = "clarify"

    conversation_history.append(f"AI: {response}")
    audio_response_path = speak_and_save(response)
    return response, audio_response_path

# Save Conversation History
def save_conversation():
    with open("conversation_history.txt", "w") as f:
        f.write("\n".join(conversation_history))
    return "Conversation history saved successfully!"

# Gradio Interface
iface = gr.Interface(
    fn=process_order,
    inputs=gr.Audio(type="filepath"),
    outputs=[gr.Textbox(label="Text Response"), gr.Audio(label="Audio Response")],
    title="Restaurant Voice Assistant",
    description="Talk to our voice assistant to place your order or ask about the menu!",
    live=True
)

if __name__ == "__main__":
    try:
        iface.launch()
    finally:
        save_conversation()