Update app.py

app.py

@@ -6,6 +6,9 @@ import gradio as gr
 from dotenv import load_dotenv
 import os
 import tempfile
+import speech_recognition as sr
+import io
+import soundfile as sf
 
 # Load environment variables
 load_dotenv()
@@ -22,6 +25,9 @@ TTS_PASSWORD = os.getenv("TTS_PASSWORD")
 TTS_VOICE = os.getenv("TTS_VOICE", "coral")
 TTS_SEED = int(os.getenv("TTS_SEED", "12345"))
 
+# إعداد Speech Recognition
+recognizer = sr.Recognizer()
+
 # التحقق من وجود المتغيرات المطلوبة
 required_env_vars = {
     "DEEPSEEK_API_KEY": API_KEY,
@@ -159,16 +165,44 @@ def format_response(response_dict):
     html += "</div>"
     return html
 
-def chat(message, history):
+def speech_to_text(audio_path):
+    """Convert speech to text using speech_recognition"""
+    try:
+        # Load audio file
+        with sr.AudioFile(audio_path) as source:
+            # Record the audio file
+            audio = recognizer.record(source)
+            # Use Google Speech Recognition
+            text = recognizer.recognize_google(audio)
+            return text
+    except Exception as e:
+        print(f"Error in speech recognition: {str(e)}")
+        return None
+
+def chat(audio, history):
     """Handle chat interactions"""
-    if not message:
-        # Generate welcome message for empty input
-        welcome = tutor.get_welcome_message()
-        audio_path = tutor.text_to_speech(welcome)[0]
-        return welcome, audio_path
+    if audio is None:
+        # Return empty response if no audio
+        return history, None
+    
+    # Convert audio to WAV format for speech recognition
+    audio_data = audio[1]  # Get the numpy array
+    sample_rate = audio[0]  # Get the sample rate
+    
+    # Save as temporary WAV file
+    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_wav:
+        sf.write(temp_wav.name, audio_data, sample_rate)
+        # Convert speech to text
+        audio_text = speech_to_text(temp_wav.name)
+    
+    # Clean up temporary file
+    os.unlink(temp_wav.name)
+    
+    if not audio_text:
+        return history, None
     
     # Get bot response
-    response = tutor.get_bot_response(message)
+    response = tutor.get_bot_response(audio_text)
     
     # Generate audio for the main response
     audio_path = tutor.text_to_speech(response["response"])[0]
@@ -176,24 +210,35 @@ def chat(message, history):
     # Format the complete response
     formatted_response = format_response(response)
     
-    return formatted_response, audio_path
+    # Update history in the correct format for gr.Chatbot
+    history = history or []
+    history.append((audio_text, formatted_response))
+    
+    return history, audio_path
+
+def show_welcome():
+    """Show welcome message on startup"""
+    welcome = tutor.get_welcome_message()
+    audio_path = tutor.text_to_speech(welcome)[0]
+    return [(None, welcome)], audio_path
 
 # Create Gradio interface
 with gr.Blocks(css="footer {display: none}") as demo:
     gr.Markdown("# 🤖 Sam - Your English Tutor")
-    gr.Markdown("Welcome to your personalized English learning session! Type your message below to start chatting.")
+    gr.Markdown("Welcome to your personalized English learning session! Click the microphone and start speaking!")
     
     chatbot = gr.Chatbot(
         show_label=False,
         height=400,
-        bubble_full_width=False,
+        type="messages"
     )
     
     with gr.Row():
-        txt = gr.Textbox(
-            show_label=False,
-            placeholder="Type your message here...",
-            container=False,
+        audio_input = gr.Audio(
+            source="microphone",
+            type="numpy",
+            label="Speak here",
+            show_label=True
         )
         audio_output = gr.Audio(
             label="Sam's Voice",
@@ -201,11 +246,25 @@ with gr.Blocks(css="footer {display: none}") as demo:
             type="filepath"
         )
     
-    txt.submit(chat, [txt, chatbot], [chatbot, audio_output])
-    txt.submit(lambda: "", [], [txt])  # Clear textbox after submit
+    # Handle audio input
+    audio_input.stop_recording(
+        fn=chat,
+        inputs=[audio_input, chatbot],
+        outputs=[chatbot, audio_output],
+        queue=False
+    )
+    
+    # Show welcome message on page load
+    demo.load_event(
+        fn=show_welcome,
+        inputs=None,
+        outputs=[chatbot, audio_output]
+    )
 
 # Launch the interface
 if __name__ == "__main__":
-    # Generate welcome message at startup
-    welcome = tutor.get_welcome_message()
-    demo.launch()
+    demo.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=False
+    )