Update app.py

app.py

@@ -1,90 +1,70 @@
 import gradio as gr
 import pandas as pd
-import asyncio
-from pydub import AudioSegment
-import numpy as np
 import edge_tts
+import asyncio
+import tempfile
 
-# Load the menu data
+# Load menu data
 def load_menu():
+    menu_file = "menu.xlsx"
     try:
-        menu_data = pd.read_excel("menu.xlsx")
-        return menu_data
+        return pd.read_excel(menu_file)
     except Exception as e:
-        raise ValueError("Menu file not found or unreadable!")
+        raise ValueError(f"Error loading menu file: {e}")
 
-# Generate menu details
-def get_menu_details():
+# Filter menu items based on preference
+def filter_menu(preference):
     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(", ")
+    menu_details = ""
+    for _, item in menu_data.iterrows():
+        menu_details += f"{item['Dish Name']} - ${item['Price ($)']}. "
+    return menu_details
 
-# 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()
-    transcription = transcribe_audio(audio_path)
+# Text-to-speech response
+def generate_tts_response(text):
+    communicate = edge_tts.Communicate(text)
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_file:
+        tmp_path = tmp_file.name
+        asyncio.run(communicate.save(tmp_path))
+    return tmp_path
 
-    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?"
+# Process user audio input and respond
+def process_audio(audio_file, preference):
+    if "menu" in preference.lower():
+        menu_details = filter_menu("All")
+        response_text = f"Here are the menu details: {menu_details}"
+    elif "veg samosa" in preference.lower():
+        response_text = "Order confirmed for Veg Samosa. It has been added to your cart."
     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:
-            response_text = f"Sorry, {transcription} is not on the menu."
+        response_text = "I'm sorry, I didn't understand that. Please try again."
 
-    audio_response = await generate_audio(response_text)
-    return audio_response, cart, response_text
+    tts_path = generate_tts_response(response_text)
+    return tts_path, response_text
 
-# Transcribe audio input (stubbed for simplicity)
-def transcribe_audio(audio_path):
-    # Mock transcription for now
-    return "veg samosa"  # Replace this with the transcription from an audio processing library.
-
-# Gradio UI setup
+# Build Gradio app
 def app():
-    cart = []
-
     with gr.Blocks() as demo:
-        gr.Markdown("### Welcome to the Menu")
-        with gr.Row():
-            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="")
+        gr.Markdown("# Welcome to the Menu")
 
-        # 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
+        # Audio input and output
+        audio_input = gr.Audio(label="Speak your preference or order", type="filepath")
+        audio_output = gr.Audio(label="Assistant Response", autoplay=True)
 
-        demo.load(lambda: asyncio.run(start_greeting()), inputs=[], outputs=assistant_response)
+        # Textbox for transcription
+        transcription = gr.Textbox(label="Transcription", placeholder="Detected text will appear here")
 
-        # Handle voice input
+        # Cart output
+        cart_output = gr.Textbox(label="Cart", placeholder="Your cart details will appear here")
+
+        # Audio processing
         audio_input.change(
-            lambda audio_path: asyncio.run(handle_input(audio_path, cart)),
-            inputs=[audio_input],
-            outputs=[assistant_response, cart_output, status],
+            process_audio, 
+            inputs=[audio_input, transcription], 
+            outputs=[audio_output, transcription]
         )
 
     return demo
 
-
 if __name__ == "__main__":
-    app().launch()
+    demo = app()
+    demo.launch()