Update app.py

app.py

@@ -1,137 +1,90 @@
 import gradio as gr
 import pandas as pd
 import asyncio
-import tempfile
-from edge_tts import Communicate
+from pydub import AudioSegment
+import numpy as np
+import edge_tts
 
-# Global cart to store ordered items
-cart = []
-order_processing = True  # Global flag to enable/disable order processing
-
-# Load Menu Data
+# Load the menu data
 def load_menu():
-    menu_file = "menu.xlsx"  # Ensure this file exists in the same directory
     try:
-        return pd.read_excel(menu_file)
+        menu_data = pd.read_excel("menu.xlsx")
+        return menu_data
     except Exception as e:
-        raise ValueError(f"Error loading menu file: {e}")
-
-# Generate Text-to-Speech Response
-async def generate_tts_response(text):
-    communicate = Communicate(text, "en-US-JennyNeural")
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_file:
-        await communicate.save(tmp_file.name)
-        return tmp_file.name
+        raise ValueError("Menu file not found or unreadable!")
 
-# Search for Dish
-def search_dish(dish_name):
+# Generate menu details
+def get_menu_details():
+    menu_data = load_menu()
+    details = "Here is the menu: "
+    for _, row in menu_data.iterrows():
+        details += f"{row['Dish Name']} for ${row['Price ($)']}, "
+    return details.strip(", ")
+
+# Text-to-Speech Conversion
+async def generate_audio(text):
+    communicate = edge_tts.Communicate(text, "en-US")
+    with open("response_audio.mp3", "wb") as audio_file:
+        await communicate.save(audio_file.name)
+    return "response_audio.mp3"
+
+# Handle user input and provide a response
+async def handle_input(audio_path, cart):
     menu_data = load_menu()
-    dish = menu_data[menu_data["Dish Name"].str.contains(dish_name, case=False, na=False)]
-    return dish
-
-# Handle Initial Greeting
-async def initial_greeting():
-    greeting_text = "Hello, welcome to Biryani Hub. Can you please tell me what you want?"
-    audio_response = await generate_tts_response(greeting_text)
-    return audio_response, "Hello! Welcome to Biryani Hub."
-
-# Handle Voice Commands
-async def handle_voice_command(audio_path):
-    global order_processing
     transcription = transcribe_audio(audio_path)
-    transcription = transcription.lower()
-
-    if "stop" in transcription:
-        order_processing = False
-        stop_text = "Order processing has been stopped."
-        audio_response = await generate_tts_response(stop_text)
-        return audio_response, "Order processing stopped.", "Your cart is safe. You can resume ordering later."
-
-    if not order_processing:
-        stop_text = "Order processing is currently stopped. Please say 'start' to resume."
-        audio_response = await generate_tts_response(stop_text)
-        return audio_response, "Order processing stopped.", stop_text
-
-    if "start" in transcription:
-        order_processing = True
-        start_text = "Order processing has been resumed."
-        audio_response = await generate_tts_response(start_text)
-        return audio_response, "Order processing resumed.", "Please proceed with your orders."
-
-    if transcription.startswith("order"):
-        dish_name = transcription.replace("order", "").strip()
-        dish = search_dish(dish_name)
-
-        if not dish.empty:
-            # Add to cart
-            item_name = dish.iloc[0]["Dish Name"]
-            item_price = dish.iloc[0]["Price ($)"]
-            cart.append({"name": item_name, "price": item_price})
 
-            confirmation_text = f"{item_name} is available for ${item_price}. Order confirmed and added to cart."
-            audio_response = await generate_tts_response(confirmation_text)
-            return audio_response, "Order Confirmed", confirmation_text
+    if transcription.lower() == "menu details":
+        response_text = get_menu_details()
+    elif transcription.lower() == "stop":
+        response_text = "Stopping interaction. Let me know when you're ready."
+    elif transcription.lower() == "start":
+        response_text = "Resuming interaction. What can I do for you?"
+    else:
+        item = menu_data[menu_data["Dish Name"].str.lower() == transcription.lower()]
+        if not item.empty:
+            dish_name = item.iloc[0]["Dish Name"]
+            price = item.iloc[0]["Price ($)"]
+            response_text = f"{dish_name} is available for ${price}. Adding it to your cart."
+            cart.append(f"{dish_name} - ${price}")
         else:
-            unavailable_text = f"Sorry, {dish_name} is not available in the menu."
-            audio_response = await generate_tts_response(unavailable_text)
-            return audio_response, "Order Unavailable", unavailable_text
-
-    if "menu details" in transcription:
-        menu_data = load_menu()
-        menu_items = "Here are the menu items: " + ", ".join(menu_data["Dish Name"].tolist())
-        audio_response = await generate_tts_response(menu_items)
-        return audio_response, "Menu Details", menu_items
+            response_text = f"Sorry, {transcription} is not on the menu."
 
-    menu_text = "I didn't understand your request. Please say a valid command."
-    audio_response = await generate_tts_response(menu_text)
-    return audio_response, "Command not recognized", menu_text
+    audio_response = await generate_audio(response_text)
+    return audio_response, cart, response_text
 
-# Transcribe Audio Placeholder
+# Transcribe audio input (stubbed for simplicity)
 def transcribe_audio(audio_path):
-    # Replace with actual transcription logic (e.g., Whisper API or Google Speech-to-Text)
-    return "order veg samosa"  # Replace this with the transcription logic
+    # Mock transcription for now
+    return "veg samosa"  # Replace this with the transcription from an audio processing library.
 
-# Update Cart Content
-def update_cart():
-    if cart:
-        cart_content = "\n".join([f"{item['name']} - ${item['price']}" for item in cart])
-    else:
-        cart_content = "Your cart is empty."
-    return cart_content
-
-# Gradio App
+# Gradio UI setup
 def app():
+    cart = []
+
     with gr.Blocks() as demo:
+        gr.Markdown("### Welcome to the Menu")
         with gr.Row():
-            gr.Markdown("## Welcome to the Menu")
+            audio_input = gr.Audio(label="Speak your preference or order", source="microphone", type="filepath")
+            assistant_response = gr.Audio(label="Assistant Response")
+            cart_output = gr.Textbox(label="Cart", value="")
+            status = gr.Textbox(label="Status", value="")
 
-        with gr.Row():
-            greet_button = gr.Button("Click to Start")
-            audio_input = gr.Audio(label="Speak your preference or order", type="filepath")
-            tts_output = gr.Audio(label="Assistant Response", autoplay=True)
-            cart_output = gr.Textbox(label="Cart", value="Your cart is empty.", interactive=False)
-            status_output = gr.Textbox(label="Status", value="Order processing active.", interactive=False)
+        # Initialize with a greeting message
+        async def start_greeting():
+            audio_response = await generate_audio("Hello, welcome to Biryani Hub. Can you please tell me what you want?")
+            return audio_response
 
-        greet_button.click(
-            initial_greeting,
-            inputs=[],
-            outputs=[tts_output, status_output]
-        )
+        demo.load(lambda: asyncio.run(start_greeting()), inputs=[], outputs=assistant_response)
 
+        # Handle voice input
         audio_input.change(
-            handle_voice_command,
+            lambda audio_path: asyncio.run(handle_input(audio_path, cart)),
             inputs=[audio_input],
-            outputs=[tts_output, status_output, cart_output],
-        )
-
-        gr.Button("View Cart").click(
-            update_cart,
-            inputs=[],
-            outputs=cart_output
+            outputs=[assistant_response, cart_output, status],
         )
 
     return demo
 
+
 if __name__ == "__main__":
-    demo = app()
-    demo.launch()
+    app().launch()