Update app.py

app.py

@@ -61,25 +61,6 @@ def text_to_speech(voice_id, text, session_id):
         audio_file.write(response.content)
     return audio_file_path
 
-def save_uploaded_audio(audio_file, session_id):
-    if audio_file is None:
-        return None
-    
-    # If audio_file is already a path, just copy it
-    if isinstance(audio_file, str):
-        ext = os.path.splitext(audio_file)[1]
-        if not ext:
-            ext = '.mp3'
-        output_path = f'temp_voice_{session_id}{ext}'
-        
-        # Copy the file to our temporary location
-        with open(audio_file, 'rb') as source:
-            with open(output_path, 'wb') as dest:
-                dest.write(source.read())
-        return output_path
-    
-    return None
-
 def upload_file(file_path):
     with open(file_path, 'rb') as file:
         files = {'fileToUpload': (os.path.basename(file_path), file)}
@@ -154,20 +135,22 @@ def combine_audio_video(video_path, audio_path, output_path):
 
     subprocess.run(cmd, check=True)
 
-def process_video(voice, model, text, audio_file, input_type, progress=gr.Progress()):
+def process_video(voice_id, model, text, audio_file, input_type, progress=gr.Progress()):
     session_id = str(uuid.uuid4())
     
-    # Handle audio based on input type
+    # Handle audio source based on input type
     if input_type == "text":
         progress(0, desc="Generating speech...")
-        audio_path = text_to_speech(voice, text, session_id)
+        audio_path = text_to_speech(voice_id, text, session_id)
         if not audio_path:
             return None, "Failed to generate speech audio."
     else:  # audio upload
         progress(0, desc="Processing uploaded audio...")
-        audio_path = save_uploaded_audio(audio_file, session_id)
-        if not audio_path:
-            return None, "Failed to process uploaded audio."
+        # Save uploaded audio to temporary file
+        temp_audio_path = f'temp_voice_{session_id}.mp3'
+        with open(temp_audio_path, 'wb') as f:
+            f.write(open(audio_file, 'rb').read())
+        audio_path = temp_audio_path
     
     progress(0.2, desc="Processing video...")
     video_path = os.path.join("models", model)
@@ -222,37 +205,35 @@ def create_interface():
     
     with gr.Blocks() as app:
         gr.Markdown("# JSON Train")
-        
-        input_type = gr.Radio(
-            choices=["text", "audio"],
-            label="Input Type",
-            value="text"
-        )
-        
-        with gr.Column():
-            # Text-to-speech inputs
-            with gr.Column(visible=True) as text_inputs:
-                voice_dropdown = gr.Dropdown(
-                    choices=[v[0] for v in voices],
-                    label="Select Voice",
-                    value=voices[0][0] if voices else None
+        with gr.Row():
+            with gr.Column():
+                input_type = gr.Radio(
+                    choices=["text", "audio"],
+                    label="Input Type",
+                    value="text"
                 )
-                text_input = gr.Textbox(label="Enter text", lines=3)
-            
-            # Audio upload input
-            with gr.Column(visible=False) as audio_inputs:
-                audio_upload = gr.Audio(label="Upload Audio", type="filepath")
+                
+                with gr.Column(visible=True) as text_inputs:
+                    voice_dropdown = gr.Dropdown(
+                        choices=[v[0] for v in voices],
+                        label="Select Voice",
+                        value=voices[0][0] if voices else None
+                    )
+                    text_input = gr.Textbox(label="Enter text", lines=3)
+                
+                with gr.Column(visible=False) as audio_inputs:
+                    audio_upload = gr.Audio(label="Upload Audio", type="filepath")
+                
+                model_dropdown = gr.Dropdown(
+                    choices=models,
+                    label="Select Video Model",
+                    value=models[0] if models else None
+                )
+                generate_btn = gr.Button("Generate Video")
             
-            model_dropdown = gr.Dropdown(
-                choices=models,
-                label="Select Video Model",
-                value=models[0] if models else None
-            )
-            generate_btn = gr.Button("Generate Video")
-        
-        with gr.Column():
-            video_output = gr.Video(label="Generated Video")
-            status_output = gr.Textbox(label="Status", interactive=False)
+            with gr.Column():
+                video_output = gr.Video(label="Generated Video")
+                status_output = gr.Textbox(label="Status", interactive=False)
         
         def toggle_inputs(input_type):
             return (
@@ -267,9 +248,13 @@ def create_interface():
         )
         
         def on_generate(voice_name, model_name, text, audio_file, input_type):
-            voice_id = next((v[1] for v in voices if v[0] == voice_name), None)
-            if input_type == "text" and not voice_id:
-                return None, "Invalid voice selected."
+            if input_type == "text":
+                voice_id = next((v[1] for v in voices if v[0] == voice_name), None)
+                if not voice_id:
+                    return None, "Invalid voice selected."
+            else:
+                voice_id = None  # Not needed for audio upload
+                
             return process_video(voice_id, model_name, text, audio_file, input_type)
         
         generate_btn.click(