Update app.py

app.py

@@ -714,7 +714,7 @@ def generate_docx(data: dict, output_path: str) -> str:
 def process_audio(audio_file):
     """Main function to process audio and generate Word document."""
     if audio_file is None:
-        return None, "Please upload an audio file.", None
+        return None, "Please record or upload an audio file.", None
     
     api_key = os.environ.get("GEMINI_API_KEY")
     if not api_key:
@@ -750,6 +750,48 @@ def process_audio(audio_file):
     except Exception as e:
         return None, f"Error: {str(e)}", None
 
+
+def on_recording_stop(audio_data):
+    """
+    Called when recording stops. Automatically triggers processing.
+    audio_data is a tuple of (sample_rate, audio_array) from microphone recording.
+    """
+    if audio_data is None:
+        return None, "No audio recorded.", None
+    
+    # Save the recorded audio to a temporary file
+    import numpy as np
+    from scipy.io import wavfile
+    
+    sample_rate, audio_array = audio_data
+    
+    # Create temporary wav file
+    temp_dir = tempfile.gettempdir()
+    temp_path = os.path.join(temp_dir, f"recording_{int(time.time())}.wav")
+    
+    # Ensure audio is in the right format
+    if audio_array.dtype != np.int16:
+        # Normalize and convert to int16
+        if audio_array.dtype == np.float32 or audio_array.dtype == np.float64:
+            audio_array = (audio_array * 32767).astype(np.int16)
+        else:
+            audio_array = audio_array.astype(np.int16)
+    
+    wavfile.write(temp_path, sample_rate, audio_array)
+    
+    # Process the audio
+    docx_file, status, json_data = process_audio(temp_path)
+    
+    return docx_file, status, json_data
+
+
+def process_uploaded_file(audio_file):
+    """Process an uploaded audio file."""
+    if audio_file is None:
+        return None, "Please upload an audio file.", None
+    
+    return process_audio(audio_file)
+
 # ============================================================================
 # GRADIO INTERFACE
 # ============================================================================
@@ -767,35 +809,66 @@ custom_theme = gr.themes.Base(
     block_title_text_color="#111827",
 )
 
-with gr.Blocks(title="Advance Care Planning", theme=custom_theme) as demo:
+with gr.Blocks(title="Advance Care Planning") as demo:
     gr.Markdown("""
     # Advance Care Planning
     
-    Upload an audio recording of a planning conversation to generate a structured Word document summary report.
+    Record or upload an audio conversation to generate a structured Word document summary report.
     """)
     
-    with gr.Row():
-        with gr.Column(scale=1):
-            audio_input = gr.Audio(
-                label="Upload Audio Recording",
-                type="filepath",
-                sources=["upload"]
-            )
+    with gr.Tabs():
+        with gr.TabItem("Record Audio"):
+            gr.Markdown("""
+            **Instructions:** Click the microphone button to start recording. Click again to stop.
+            The recording will be automatically analyzed when you stop.
+            """)
             
-            submit_btn = gr.Button("Analyze & Generate Word Doc", variant="primary")
+            with gr.Row():
+                with gr.Column(scale=1):
+                    audio_recorder = gr.Audio(
+                        label="Recording",
+                        sources=["microphone"],
+                        type="numpy",
+                        interactive=True
+                    )
+                
+                with gr.Column(scale=1):
+                    record_status = gr.Textbox(label="Status", interactive=False)
+                    record_docx_output = gr.File(label="Download Word Document")
+            
+            with gr.Accordion("View Extracted Data (JSON)", open=False):
+                record_json_output = gr.Code(label="Extracted Data", language="json")
+            
+            # Auto-process when recording stops
+            audio_recorder.stop_recording(
+                fn=on_recording_stop,
+                inputs=[audio_recorder],
+                outputs=[record_docx_output, record_status, record_json_output]
+            )
         
-        with gr.Column(scale=1):
-            status_output = gr.Textbox(label="Status", interactive=False)
-            docx_output = gr.File(label="Download Word Document")
-    
-    with gr.Accordion("View Extracted Data (JSON)", open=False):
-        json_output = gr.Code(label="Extracted Data", language="json")
-    
-    submit_btn.click(
-        fn=process_audio,
-        inputs=[audio_input],
-        outputs=[docx_output, status_output, json_output]
-    )
+        with gr.TabItem("Upload Audio"):
+            with gr.Row():
+                with gr.Column(scale=1):
+                    audio_upload = gr.Audio(
+                        label="Upload Audio Recording",
+                        type="filepath",
+                        sources=["upload"]
+                    )
+                    
+                    upload_btn = gr.Button("Analyze & Generate Word Doc", variant="primary")
+                
+                with gr.Column(scale=1):
+                    upload_status = gr.Textbox(label="Status", interactive=False)
+                    upload_docx_output = gr.File(label="Download Word Document")
+            
+            with gr.Accordion("View Extracted Data (JSON)", open=False):
+                upload_json_output = gr.Code(label="Extracted Data", language="json")
+            
+            upload_btn.click(
+                fn=process_uploaded_file,
+                inputs=[audio_upload],
+                outputs=[upload_docx_output, upload_status, upload_json_output]
+            )
     
     gr.Markdown("""
     ---
@@ -805,4 +878,4 @@ with gr.Blocks(title="Advance Care Planning", theme=custom_theme) as demo:
     """)
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(theme=custom_theme)