Update app.py

app.py

@@ -1,36 +1,86 @@
 import gradio as gr
-from tts_module import text_to_speech  # Módulo para TTS
-from video_module import generate_video  # Módulo para generar video
-from music_module import adjust_background_music  # Módulo para manejar música
-
-def process_input(text, txt_file, mp3_file, prompt):
-    # Paso 1: Obtener el texto final
-    if text.strip():  # Prioridad al texto escrito directamente
-        final_text = text
-    elif txt_file is not None:  # Luego al archivo .txt
-        final_text = txt_file.decode("utf-8")
-    elif prompt.strip():  # Finalmente, usar el prompt para generar texto
-        # Aquí iría la llamada al modelo gpt2 para generar el guion
-        final_text = "Generated text from prompt"  # Placeholder
-    else:
-        return "No input provided", None
-
-    # Paso 2: Convertir texto a voz con edge_tts
-    audio_file = text_to_speech(final_text)
-
-    # Paso 3: Generar el video
-    video_file = generate_video(final_text)
-
-    # Paso 4: Ajustar la música de fondo
-    if mp3_file is not None:
-        music_clip = adjust_background_music(video_file.duration, mp3_file.name)
-    else:
-        music_clip = None
-
-    # Paso 5: Combinar todos los elementos
-    final_video = combine_audio_video(audio_file, video_file, music_clip)
-
-    return final_video
+from tts_module import list_voices, text_to_speech
+from pexels_api import search_pexels  # Asegúrate de que este archivo esté correctamente configurado
+from moviepy.editor import AudioFileClip, VideoFileClip, CompositeAudioClip
+import asyncio
+import os
+import requests
+import tempfile
+
+# Obtener lista de voces disponibles
+async def get_voices():
+    voices = await list_voices()
+    return list(voices.keys())
+
+# Ajustar música de fondo
+def adjust_background_music(video_duration, music_file):
+    from moviepy.editor import AudioFileClip, concatenate_audioclips
+    music = AudioFileClip(music_file)
+    if music.duration < video_duration:
+        repetitions = int(video_duration / music.duration) + 1
+        music_clips = [music] * repetitions
+        music = concatenate_audioclips(music_clips)
+    if music.duration > video_duration:
+        music = music.subclip(0, video_duration)
+    music = music.volumex(0.2)  # Reducir volumen al 20%
+    return music
+
+# Combinar audio, video y música
+def combine_audio_video(audio_file, video_clip, music_clip=None):
+    audio_clip = AudioFileClip(audio_file)
+    final_clip = video_clip.set_audio(audio_clip)
+
+    if music_clip:
+        final_clip = final_clip.set_audio(CompositeAudioClip([audio_clip, music_clip]))
+    
+    output_path = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4").name
+    final_clip.write_videofile(output_path, codec="libx264", audio_codec="aac")
+    return output_path
+
+# Función principal
+def process_input(text, txt_file, mp3_file, prompt, selected_voice):
+    try:
+        # Paso 1: Obtener el texto final
+        if text.strip():  # Prioridad al texto escrito directamente
+            final_text = text
+        elif txt_file is not None:  # Luego al archivo .txt
+            final_text = txt_file.decode("utf-8")
+        elif prompt.strip():  # Finalmente, usar el prompt para generar texto
+            final_text = "Generated text from prompt"  # Placeholder
+        else:
+            return "No input provided", None
+
+        # Paso 2: Convertir texto a voz con edge_tts
+        audio_file = asyncio.run(text_to_speech(final_text, selected_voice))
+
+        # Paso 3: Buscar videos en Pexels
+        query = final_text.split(".")[0]  # Usar la primera oración como consulta
+        try:
+            video_links = search_pexels(query, num_results=1)
+            if not video_links:
+                return "No se encontraron videos en Pexels", None
+        except Exception as e:
+            return f"Error al buscar videos en Pexels: {e}", None
+
+        # Descargar el primer video encontrado
+        video_response = requests.get(video_links[0])
+        with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as tmp_video:
+            tmp_video.write(video_response.content)
+            video_clip = VideoFileClip(tmp_video.name)
+
+        # Paso 4: Ajustar la música de fondo
+        if mp3_file is not None:
+            music_clip = adjust_background_music(video_clip.duration, mp3_file.name)
+        else:
+            music_clip = None
+
+        # Paso 5: Combinar todos los elementos
+        final_video = combine_audio_video(audio_file, video_clip, music_clip)
+
+        return final_video
+
+    except Exception as e:
+        return f"Error durante el procesamiento: {e}", None
 
 # Interfaz Gradio
 with gr.Blocks() as demo:
@@ -41,10 +91,13 @@ with gr.Blocks() as demo:
             txt_file_input = gr.File(label="Or upload a .txt file", file_types=[".txt"])
             mp3_file_input = gr.File(label="Upload background music (.mp3)", file_types=[".mp3"])
             prompt_input = gr.Textbox(label="Or enter a prompt to generate text")
+            voices = asyncio.run(get_voices())
+            voice_dropdown = gr.Dropdown(choices=voices, label="Select Voice")
+
         with gr.Column():
             output_video = gr.Video(label="Generated Video")
 
     btn = gr.Button("Generate Video")
-    btn.click(process_input, inputs=[text_input, txt_file_input, mp3_file_input, prompt_input], outputs=output_video)
+    btn.click(process_input, inputs=[text_input, txt_file_input, mp3_file_input, prompt_input, voice_dropdown], outputs=output_video)
 
 demo.launch()