Update app.py

app.py

@@ -2,33 +2,37 @@ import gradio as gr
 import os
 import subprocess
 import ssl
-from moviepy.editor import VideoFileClip, AudioFileClip, TextClip, CompositeVideoClip, concatenate_videoclips
+import time
+from moviepy.editor import VideoFileClip, AudioFileClip, TextClip, CompositeVideoClip
 from transformers import pipeline
-import torchaudio
-import librosa
-import numpy as np
 import ffmpeg
 
 # Configuração para evitar erro de SSL
 ssl._create_default_https_context = ssl._create_unverified_context
 
+# Função para atualizar a barra de progresso
+def update_progress(current, total, progress=gr.Progress()):
+    progress(current / total)
+
 # Função para extrair áudio do vídeo
-def extract_audio(video_path):
+def extract_audio(video_path, progress):
     video = VideoFileClip(video_path)
-    if video.audio is None:
-        raise ValueError("O vídeo não contém áudio.")
     audio_path = "temp_audio.wav"
+    update_progress(10, 100, progress)
     video.audio.write_audiofile(audio_path, codec='pcm_s16le')
+    update_progress(20, 100, progress)
     return audio_path
 
 # Função para transcrever o áudio e gerar legendas
-def transcribe_audio(audio_path, language):
+def transcribe_audio(audio_path, language, progress):
     transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-large-v2")
+    update_progress(30, 100, progress)
     result = transcriber(audio_path, return_timestamps=True)
+    update_progress(40, 100, progress)
     return result['text'], result['chunks']
 
-# Função para clonar a voz e gerar dublagem usando VITS
-def clone_voice_and_generate_dub(text, language, audio_path):
+# Função para clonar a voz e gerar dublagem
+def clone_voice_and_generate_dub(text, language, audio_path, progress):
     cloned_voice_path = "dub_audio.wav"
     command = [
         "python", "vits_inference.py", 
@@ -37,33 +41,43 @@ def clone_voice_and_generate_dub(text, language, audio_path):
         "--output_audio", cloned_voice_path, 
         "--language", language
     ]
-    subprocess.run(command, check=True)
+    update_progress(50, 100, progress)
+    subprocess.run(command)
+    update_progress(70, 100, progress)
     return cloned_voice_path
 
 # Função para adicionar legendas ao vídeo
-def add_subtitles(video_clip, subtitles, primary_lang, secondary_lang=None):
+def add_subtitles(video_clip, subtitles, primary_lang, secondary_lang, progress):
     subtitle_clips = []
     for chunk in subtitles:
         start, end, text = chunk['timestamp'][0], chunk['timestamp'][1], chunk['text']
-        subtitle = TextClip(text, fontsize=30, color='white', bg_color='black', size=(video_clip.w, 50))
-        subtitle = subtitle.set_start(start).set_duration(end - start).set_position(('center', 'bottom'))
+        subtitle = TextClip(text, fontsize=30, color='white', bg_color='black', size=(video_clip.w, 50)).set_start(start).set_duration(end - start).set_position(('center', 'bottom'))
         subtitle_clips.append(subtitle)
+    update_progress(80, 100, progress)
     return CompositeVideoClip([video_clip] + subtitle_clips)
 
-# Função para processar o vídeo final
-def process_video(video_path, format_choice, add_subtitles_option, dual_subtitles, primary_lang, secondary_lang, dub_language):
-    audio_path = extract_audio(video_path)
-    text, subtitles = transcribe_audio(audio_path, primary_lang)
-    dub_audio_path = clone_voice_and_generate_dub(text, dub_language, audio_path)
+# Função para processar o vídeo
+def process_video(video_path, format_choice, add_subtitles_option, dual_subtitles, primary_lang, secondary_lang, dub_language, progress=gr.Progress()):
+    start_time = time.time()
+    
+    audio_path = extract_audio(video_path, progress)
+    text, subtitles = transcribe_audio(audio_path, primary_lang, progress)
+    dub_audio_path = clone_voice_and_generate_dub(text, dub_language, audio_path, progress)
     
     video = VideoFileClip(video_path).set_audio(AudioFileClip(dub_audio_path))
     
     if add_subtitles_option:
-        video = add_subtitles(video, subtitles, primary_lang, secondary_lang if dual_subtitles else None)
+        video = add_subtitles(video, subtitles, primary_lang, secondary_lang if dual_subtitles else None, progress)
     
     final_video_path = "final_video.mp4"
+    update_progress(90, 100, progress)
     video.write_videofile(final_video_path, codec='libx264', fps=30)
-    return final_video_path
+    update_progress(100, 100, progress)
+    
+    end_time = time.time()
+    estimated_time = round(end_time - start_time, 2)
+    
+    return final_video_path, f"✅ Processamento concluído! Tempo estimado: {estimated_time} segundos."
 
 # Interface Gradio
 iface = gr.Interface(
@@ -73,11 +87,11 @@ iface = gr.Interface(
         gr.Radio(["TikTok/Kwai (9:16)", "YouTube (16:9)"], label="Formato de Saída"),
         gr.Checkbox(label="Adicionar Legendas?"),
         gr.Checkbox(label="Usar Legenda Dupla?"),
-        gr.Dropdown(["en", "es", "fr", "de", "pt"], label="Idioma Principal das Legendas"),
+        gr.Dropdown(["en", "es", "fr", "de", "pt"], label="Idioma Principal das Legendas e Dublagem"),
         gr.Dropdown(["en", "es", "fr", "de", "pt"], label="Idioma Secundário da Legenda", value=None),
-        gr.Dropdown(["en", "es", "fr", "de", "pt"], label="Idioma da Dublagem")
+        gr.Dropdown(["en", "es", "fr", "de", "pt"], label="Idioma da Dublagem"),
     ],
-    outputs=gr.Video(label="Vídeo Final"),
+    outputs=[gr.Video(label="Vídeo Final"), gr.Textbox(label="Status do Processamento")],
     title="Ferramenta de Dublagem e Legendas Automáticas",
     description="Carregue um vídeo, escolha as opções desejadas e receba o vídeo dublado com legendas automáticas!"
 )