Update app.py

app.py

@@ -31,45 +31,50 @@ async def text_to_speech(text, voice, rate, pitch):
     except Exception as e:
         return None, f"Speech generation failed: {str(e)}"
 
-# Agregar música de fondo
+# Agregar música de fondo (ahora elimina el audio original)
 def add_background_music(speech_path, bg_music_path):
     speech = AudioSegment.from_file(speech_path)
     background = AudioSegment.from_file(bg_music_path) - 16  # 15% volume
     
-    # Asegurar que la música de fondo dure al menos como el speech + 3s fadeout
     if len(background) < len(speech) + 3000:
         background = background * math.ceil((len(speech)+3000)/len(background))
     
-    # Combinar audio con fadeout
     combined = speech.overlay(background[:len(speech)])
     fade_out = background[len(speech):len(speech)+3000].fade_out(3000)
     final_audio = combined + fade_out
     
     with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_file:
         final_audio.export(tmp_file.name, format="mp3")
+        # Eliminar audio original
+        if os.path.exists(speech_path):
+            os.remove(speech_path)
         return tmp_file.name
 
-# Procesar múltiples videos
+# Procesar múltiples videos (ahora elimina archivos temporales)
 def process_videos(audio_path, video_files):
-    audio_duration = AudioSegment.from_file(audio_path).duration_seconds
-    
-    # Crear video concatenado
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as concat_video:
-        # Crear lista de videos para concatenar
-        with tempfile.NamedTemporaryFile(mode="w", delete=False, suffix=".txt") as list_file:
-            list_file.write("\n".join([f"file '{v.name}'" for v in video_files]))
-            list_file.close()
-            
-            subprocess.run([
-                "ffmpeg", "-y",
-                "-f", "concat",
-                "-safe", "0",
-                "-i", list_file.name,
-                "-c", "copy",
-                concat_video.name
-            ], check=True)
+    temp_files = []
+    try:
+        audio_duration = AudioSegment.from_file(audio_path).duration_seconds
         
-        # Crear video final con loop
+        # Concatenar videos
+        with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as concat_video:
+            temp_files.append(concat_video.name)
+            
+            with tempfile.NamedTemporaryFile(mode="w", delete=False, suffix=".txt") as list_file:
+                temp_files.append(list_file.name)
+                list_file.write("\n".join([f"file '{v.name}'" for v in video_files]))
+                list_file.close()
+                
+                subprocess.run([
+                    "ffmpeg", "-y",
+                    "-f", "concat",
+                    "-safe", "0",
+                    "-i", list_file.name,
+                    "-c", "copy",
+                    concat_video.name
+                ], check=True)
+            
+        # Crear video final
         with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as final_video:
             subprocess.run([
                 "ffmpeg", "-y",
@@ -86,34 +91,49 @@ def process_videos(audio_path, video_files):
             ], check=True)
             
             return final_video.name
+            
+    finally:
+        # Eliminar archivos temporales
+        for f in temp_files:
+            if os.path.exists(f):
+                os.remove(f)
 
-# Función principal
+# Función principal (ahora elimina videos originales)
 async def tts_interface(text, voice, rate, pitch, bg_music, video_files):
-    # Generar audio principal
-    audio_path, warning = await text_to_speech(text, voice, rate, pitch)
-    if warning:
-        return None, None, gr.Warning(warning)
-    
+    temp_audio = None
     try:
+        # Generar audio principal
+        temp_audio, warning = await text_to_speech(text, voice, rate, pitch)
+        if warning:
+            return None, None, gr.Warning(warning)
+        
         # Agregar música de fondo
         if bg_music:
-            audio_path = add_background_music(audio_path, bg_music)
+            temp_audio = add_background_music(temp_audio, bg_music)
         
         # Procesar videos
+        video_path = None
         if video_files:
-            video_path = process_videos(audio_path, video_files)
-        else:
-            video_path = None
+            video_path = process_videos(temp_audio, video_files)
+            # Eliminar videos originales subidos
+            for video in video_files:
+                if hasattr(video, 'name') and os.path.exists(video.name):
+                    os.remove(video.name)
         
-        return audio_path, video_path, None
+        return temp_audio, video_path, None
     
     except Exception as e:
         return None, None, gr.Warning(f"Processing error: {str(e)}")
     finally:
-        if 'audio_path' in locals() and os.path.exists(audio_path):
-            os.remove(audio_path)
+        # Eliminar audio temporal si existe y no es la salida final
+        if temp_audio and os.path.exists(temp_audio):
+            try:
+                if video_path and temp_audio != video_path:
+                    os.remove(temp_audio)
+            except:  # Evitar errores si el archivo ya fue eliminado
+                pass
 
-# Crear interfaz
+# Crear interfaz (sin cambios)
 async def create_demo():
     voices = await get_voices()