Update app.py

app.py

@@ -1,14 +1,8 @@
 import gradio as gr
 from gtts import gTTS
-import os
 import tempfile
-import json
-import sounddevice as sd
-import queue
-import threading
-from scipy.io.wavfile import write
+import os
 import speech_recognition as sr
-import subprocess
 
 # Store cart in a temporary storage
 cart = []
@@ -22,9 +16,6 @@ menu_items = {
     "Soda": 2.49
 }
 
-# To manage playback control
-playback_control = {"stop": False, "pause": False}
-
 def generate_voice_response(text):
     tts = gTTS(text)
     temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
@@ -32,96 +23,69 @@ def generate_voice_response(text):
     tts.save(temp_file.name)
     return temp_file.name
 
-def play_audio(audio_path):
+def process_audio(audio_path):
+    global cart
+    recognizer = sr.Recognizer()
+    response = ""
+
     try:
-        subprocess.run(["ffplay", "-nodisp", "-autoexit", audio_path], check=True)
-    except Exception as e:
-        print("Error playing audio:", e)
+        with sr.AudioFile(audio_path) as source:
+            audio = recognizer.record(source)
+            input_text = recognizer.recognize_google(audio)
+            print("User said:", input_text)
 
-def calculate_total(cart):
-    return sum(menu_items[item] for item in cart)
+            if "menu" in input_text.lower():
+                response = "Here is our menu:\n"
+                for item, price in menu_items.items():
+                    response += f"{item}: ${price:.2f}\n"
+                response += "\nWhat would you like to add to your cart?"
 
-# Audio recording setup
-q = queue.Queue()
-def audio_callback(indata, frames, time, status):
-    if status:
-        print(status)
-    q.put(indata.copy())
+            elif any(item.lower() in input_text.lower() for item in menu_items):
+                for item in menu_items:
+                    if item.lower() in input_text.lower():
+                        cart.append(item)
+                        total = sum(menu_items[cart_item] for cart_item in cart)
+                        response = f"{item} has been added to your cart. Your current cart includes:\n"
+                        for cart_item in cart:
+                            response += f"- {cart_item}: ${menu_items[cart_item]:.2f}\n"
+                        response += f"\nTotal: ${total:.2f}. Would you like to add anything else?"
+                        break
 
-def record_audio():
-    fs = 16000  # Sample rate
-    seconds = 5  # Duration
-    print("Recording...")
-    audio_data = []
-    with sd.InputStream(callback=audio_callback, samplerate=fs, channels=1):
-        for _ in range(int(fs * seconds / 1024)):
-            audio_data.append(q.get())
-    print("Recording finished.")
-    return b"".join(audio_data)
+            elif "final order" in input_text.lower() or "submit order" in input_text.lower():
+                if cart:
+                    total = sum(menu_items[cart_item] for cart_item in cart)
+                    response = "Your final order includes:\n"
+                    for item in cart:
+                        response += f"- {item}: ${menu_items[item]:.2f}\n"
+                    response += f"\nTotal: ${total:.2f}.\nThank you for ordering!"
+                    cart = []  # Clear cart after finalizing order
+                else:
+                    response = "Your cart is empty. Would you like to order something?"
 
-def restaurant_voice_assistant():
-    global cart
-    recognizer = sr.Recognizer()
+            else:
+                response = "I didn’t quite catch that. Please tell me what you’d like to order."
 
-    # Continuously listen for user commands
-    while True:
-        try:
-            audio_data = record_audio()
-            with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
-                temp_audio.write(audio_data)
-                temp_audio_path = temp_audio.name
+    except sr.UnknownValueError:
+        response = "Sorry, I didn’t catch that. Could you please repeat?"
+    except Exception as e:
+        response = f"An error occurred: {str(e)}"
 
-            with sr.AudioFile(temp_audio_path) as source:
-                audio = recognizer.record(source)
-                input_text = recognizer.recognize_google(audio)
-                print("You said:", input_text)
+    audio_response_path = generate_voice_response(response)
+    return response, audio_response_path
 
-                # Process input text
-                if "menu" in input_text.lower():
-                    response = "Here is our menu:\n"
-                    for item, price in menu_items.items():
-                        response += f"{item}: ${price:.2f}\n"
-                    response += "\nWhat would you like to add to your cart?"
-                elif any(item.lower() in input_text.lower() for item in menu_items):
-                    for item in menu_items:
-                        if item.lower() in input_text.lower():
-                            cart.append(item)
-                            total = calculate_total(cart)
-                            response = f"{item} has been added to your cart. Your current cart includes:\n"
-                            for cart_item in cart:
-                                response += f"- {cart_item}: ${menu_items[cart_item]:.2f}\n"
-                            response += f"\nTotal: ${total:.2f}. Would you like to add anything else?"
-                            break
-                elif "final order" in input_text.lower() or "submit order" in input_text.lower():
-                    if cart:
-                        total = calculate_total(cart)
-                        response = "Your final order includes:\n"
-                        for item in cart:
-                            response += f"- {item}: ${menu_items[item]:.2f}\n"
-                        response += f"\nTotal: ${total:.2f}.\nThank you for ordering!"
-                        cart = []  # Clear cart after finalizing order
-                    else:
-                        response = "Your cart is empty. Would you like to order something?"
-                elif "stop" in input_text.lower():
-                    playback_control["stop"] = True
-                    response = "Audio playback stopped."
-                elif "pause" in input_text.lower():
-                    playback_control["pause"] = True
-                    response = "Audio playback paused."
-                elif "resume" in input_text.lower():
-                    playback_control["pause"] = False
-                    response = "Resuming audio playback."
-                else:
-                    response = "I didn’t quite catch that. Please tell me what you’d like to order."
+def run_assistant(audio):
+    transcription, audio_response_path = process_audio(audio)
+    return transcription, audio_response_path
 
-                # Generate and play audio response
-                audio_path = generate_voice_response(response)
-                play_audio(audio_path)
+# Gradio Interface
+with gr.Blocks() as demo:
+    gr.Markdown("# Voice-Activated Restaurant Assistant")
+    with gr.Row():
+        audio_input = gr.Audio(source="microphone", type="filepath", label="Speak Now")
+        text_output = gr.Textbox(label="Transcription")
+        audio_output = gr.Audio(label="Assistant Response")
 
-        except sr.UnknownValueError:
-            print("Sorry, I didn’t catch that. Could you please repeat?")
-        except Exception as e:
-            print("Error occurred:", e)
+    submit_button = gr.Button("Submit")
+    submit_button.click(run_assistant, inputs=[audio_input], outputs=[text_output, audio_output])
 
-# Run the assistant
-threading.Thread(target=restaurant_voice_assistant).start()
+demo.launch()