Update app.py

app.py

@@ -36,7 +36,7 @@ class WebSocketClient:
                 await self.websocket.send(f.read())
             await asyncio.gather(self.receive_messages(), self.send_audio_chunks())
         except Exception as e:
-            print(f"🔴 WebSocket Connection Failed: {e}")
+            print(f"\U0001F534 WebSocket Connection Failed: {e}")
 
     def run(self):
         asyncio.set_event_loop(self.loop)
@@ -115,7 +115,7 @@ def handle_chat(user_input, history, thread_id, image_url):
                 content = msg.content[0].text.value
                 history.append((user_input, content))
                 match = re.search(
-                    r'https://raw\.githubusercontent\.com/AndrewLORTech/surgical-pathology-manual/main/[\w\-/]*\.png',
+                    r'https://raw\\.githubusercontent\\.com/AndrewLORTech/surgical-pathology-manual/main/[\\w\\-/]*\\.png',
                     content
                 )
                 if match: image_url = match.group(0)
@@ -150,21 +150,48 @@ with gr.Blocks(theme=gr.themes.Soft()) as app:
 
             with gr.Accordion("🎤 Voice Transcription", open=False) as voice_section:
                 with gr.Row():
-                    voice_input = gr.Audio(label="Mic", streaming=True)
+                    voice_input = gr.Audio(label="🎙️ Record", streaming=True, elem_classes="record-button")
                     voice_transcript = gr.Textbox(label="Transcript", lines=2, interactive=False)
-                clear_btn = gr.Button("🧹 Clear Transcript")
+                with gr.Row():
+                    ask_btn = gr.Button("🟢 Ask", elem_id="ask-btn")
+                    clear_chat_btn = gr.Button("🧹 Clear Chat", elem_id="clear-chat-btn")
 
     # Functional bindings
     def toggle_voice(curr):
         return not curr, gr.update(visible=not curr)
 
+    def send_transcript_to_assistant(transcript, history, thread_id, image_url):
+        if not transcript.strip():
+            return gr.update(), history, thread_id, image_url
+        return handle_chat(transcript, history, thread_id, image_url)
+
+    def clear_chat_and_transcript(client_id):
+        if client_id in connections:
+            connections[client_id].transcript = ""
+        return [], "", None, None
+
     mic_toggle_btn.click(fn=toggle_voice, inputs=voice_enabled, outputs=[voice_enabled, voice_section])
     send_btn.click(fn=handle_chat,
                    inputs=[user_prompt, chat_state, thread_state, image_state],
                    outputs=[user_prompt, chat, thread_state, image_state])
     image_state.change(fn=lambda x: x, inputs=image_state, outputs=image_display)
     voice_input.stream(fn=send_audio, inputs=[voice_input, client_id], outputs=voice_transcript, stream_every=0.5)
-    clear_btn.click(fn=clear_transcript, inputs=[client_id], outputs=voice_transcript)
+    ask_btn.click(fn=send_transcript_to_assistant,
+                  inputs=[voice_transcript, chat_state, thread_state, image_state],
+                  outputs=[user_prompt, chat, thread_state, image_state])
+    clear_chat_btn.click(fn=clear_chat_and_transcript,
+                         inputs=[client_id],
+                         outputs=[chat, voice_transcript, thread_state, image_state])
     app.load(fn=create_ws, outputs=[client_id])
 
-app.launch()
+app.launch(css="""
+#ask-btn, #clear-chat-btn {
+    font-size: 16px !important;
+    padding: 10px 20px !important;
+}
+.record-button button {
+    font-size: 16px !important;
+    padding: 12px 24px !important;
+    background-color: #f2f2f2;
+}
+""")