Create app.py

app.py

@@ -0,0 +1,152 @@
+import os
+import math
+import tempfile
+from pydub import AudioSegment
+from moviepy.editor import (
+    VideoFileClip, AudioFileClip, ImageClip, 
+    concatenate_videoclips, CompositeVideoClip,
+    concatenate_audioclips
+)
+import edge_tts
+import gradio as gr
+import asyncio
+
+# CONSTANTES DE ARCHIVOS
+INTRO_VIDEO = "introvideo.mp4"
+OUTRO_VIDEO = "outrovideo.mp4"
+MUSIC_BG = "musicafondo.mp3"
+FX_SOUND = "fxsound.mp3"
+WATERMARK = "watermark.png"
+
+# Validar existencia de archivos obligatorios
+for file in [INTRO_VIDEO, OUTRO_VIDEO, MUSIC_BG, FX_SOUND, WATERMARK]:
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"Falta archivo necesario: {file}")
+
+def cortar_video(video_path, metodo="inteligente", duracion=10):
+    video = VideoFileClip(video_path)
+    if metodo == "manual":
+        return [video.subclip(i*duracion, (i+1)*duracion) 
+                for i in range(math.ceil(video.duration/duracion))]
+    
+    # Implementación básica de cortes por voz (requeriría VAD real)
+    clips = []
+    ultimo_corte = 0
+    for i in range(1, math.ceil(video.duration)):
+        if i % 5 == 0:  # Simulación de detección de pausas
+            clips.append(video.subclip(ultimo_corte, i))
+            ultimo_corte = i
+    return clips
+
+def procesar_audio(texto, voz, clips_duracion):
+    communicate = edge_tts.Communicate(texto, voz)
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp:
+        asyncio.run(communicate.save(tmp.name))
+        tts_audio = AudioFileClip(tmp.name)
+    
+    # Ajustar TTS a duración de clips
+    if tts_audio.duration < clips_duracion:
+        tts_audio = tts_audio.loop(duration=clips_duracion)
+    else:
+        tts_audio = tts_audio.subclip(0, clips_duracion)
+    
+    # Mezclar con música de fondo
+    bg_music = AudioSegment.from_mp3(MUSIC_BG)
+    if len(bg_music) < clips_duracion*1000:
+        bg_music = bg_music * math.ceil(clips_duracion*1000 / len(bg_music))
+    bg_music = bg_music[:clips_duracion*1000].fade_out(3000)
+    
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp:
+        bg_music.export(tmp.name, format="mp3")
+        bg_audio = AudioFileClip(tmp.name).volumex(0.10)
+    
+    return CompositeAudioClip([bg_audio, tts_audio.volumex(0.9)])
+
+def agregar_transiciones(clips):
+    fx_audio = AudioFileClip(FX_SOUND).set_duration(2.5)
+    transicion = ImageClip(WATERMARK).set_duration(2.5)
+    transicion = transicion.resize(height=clips[0].h).set_position(("center", 0.1))
+    
+    clips_con_fx = []
+    for i, clip in enumerate(clips):
+        # Agregar watermark
+        clip_watermarked = CompositeVideoClip([clip, transicion])
+        clips_con_fx.append(clip_watermarked)
+        
+        if i < len(clips)-1:
+            clips_con_fx.append(
+                CompositeVideoClip([transicion.set_position("center")])
+                .set_audio(fx_audio)
+            )
+    
+    return concatenate_videoclips(clips_con_fx)
+
+async def procesar_video(
+    video_input, 
+    texto_tts,
+    voz_seleccionada,
+    metodo_corte,
+    duracion_corte
+):
+    # Procesar video principal
+    clips = cortar_video(video_input, metodo_corte, duracion_corte)
+    video_editado = agregar_transiciones(clips)
+    
+    # Agregar intro/outro
+    intro = VideoFileClip(INTRO_VIDEO)
+    outro = VideoFileClip(OUTRO_VIDEO)
+    video_final = concatenate_videoclips([intro, video_editado, outro])
+    
+    # Procesar audio
+    audio_final = procesar_audio(
+        texto_tts, 
+        voz_seleccionada,
+        video_editado.duration
+    )
+    
+    # Combinar y renderizar
+    video_final = video_final.set_audio(audio_final)
+    
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as tmp:
+        video_final.write_videofile(tmp.name, codec="libx264", fps=24)
+        return tmp.name
+
+# Interfaz Gradio
+with gr.Blocks() as demo:
+    gr.Markdown("# Video Editor IA")
+    
+    with gr.Tab("Principal"):
+        video_input = gr.Video(label="Subir video")
+        texto_tts = gr.Textbox(label="Texto para TTS", lines=3)
+        voz_seleccionada = gr.Dropdown(
+            label="Seleccionar voz",
+            choices=["es-ES-AlvaroNeural", "es-MX-BeatrizNeural"]
+        )
+        procesar_btn = gr.Button("Generar Video")
+        video_output = gr.Video(label="Resultado")
+    
+    with gr.Tab("Ajustes"):
+        metodo_corte = gr.Radio(
+            ["inteligente", "manual"],
+            label="Método de cortes",
+            value="inteligente"
+        )
+        duracion_corte = gr.Slider(
+            1, 60, 10,
+            label="Segundos por corte (solo manual)"
+        )
+
+    procesar_btn.click(
+        procesar_video,
+        inputs=[
+            video_input,
+            texto_tts,
+            voz_seleccionada,
+            metodo_corte,
+            duracion_corte
+        ],
+        outputs=video_output
+    )
+
+if __name__ == "__main__":
+    demo.queue().launch()