Update app.py

app.py

@@ -379,7 +379,12 @@ def process_voice_note(audio_file, history):
     if audio_file is None:
         return "Please record or upload an audio file.", history, "", None, None
     try:
-        transcript = rag.transcribe_audio(audio_file)
+        # If audio_file is a string (filepath), open it as a file
+        if isinstance(audio_file, str):
+            with open(audio_file, "rb") as f:
+                transcript = rag.transcribe_audio(f)
+        else:
+            transcript = rag.transcribe_audio(audio_file)
         if not transcript or not str(transcript).strip():
             history.append({"role": "user", "content": "🎤 [No audio detected or transcription failed]"})
             history.append({"role": "assistant", "content": "Sorry, I couldn't understand the audio. Please try again."})
@@ -403,19 +408,16 @@ with gr.Blocks(css=custom_css, title="Document Q&A System") as demo:
     <div style='text-align:center; color:#1976D2; margin-bottom:18px;'>Upload a document, record your voice, and chat!</div>
     """)
 
-    # Define shared components at the top
     chatbot = gr.Chatbot(height=400, elem_classes="gradio-chatbot", label=None, type="messages")
     file_output = gr.Textbox(label="Upload Status", interactive=False, elem_classes="textbox")
     question = gr.Textbox(label="Type your question and press Enter", placeholder="Ask a question about your document...", elem_classes="textbox")
     audio_input = gr.Audio(type="filepath", label="Record or Upload Audio", elem_classes="gradio-audio", visible=False)
-    audio_status = gr.Textbox(label="Audio Status", interactive=False, elem_classes="status-text", visible=False)
     tts_output = gr.Audio(label="Assistant Voice Reply", interactive=False, visible=False)
 
     with gr.Row():
-        # Controls on the top left
-        with gr.Column(scale=1, min_width=320):
+        with gr.Column(scale=1, min_width=350):
             with gr.Group(elem_classes="compact-box"):
-                gr.Markdown("<div class='section-title'>Upload & Voice</div>")
+                gr.Markdown("<div class='section-title'>Document Q&A Controls</div>")
                 file_input = gr.File(label="Upload Document", file_types=[".pdf", ".txt", ".doc", ".docx"], file_count="single", type="binary", elem_classes="upload-btn")
                 mic_btn = gr.Button("🎤 Record Voice", elem_classes="audio-btn")
                 audio_input
@@ -435,8 +437,7 @@ with gr.Blocks(css=custom_css, title="Document Q&A System") as demo:
                 send_voice_btn.click(process_voice_note, [audio_input, chatbot], [file_output, chatbot, question, audio_input, tts_output])
                 send_voice_btn.click(hide_audio, None, [audio_input, send_voice_btn])
                 tts_output
-        # Chatbot/chat area to the right
-        with gr.Column(scale=3, min_width=400):
+        with gr.Column(scale=3, min_width=500):
             with gr.Group(elem_classes="compact-box"):
                 chatbot
                 with gr.Row():