Update app.py
Browse files
app.py
CHANGED
|
@@ -1,69 +1,96 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
import os
|
| 3 |
import subprocess
|
| 4 |
-
from huggingface_hub import HfApi, hf_hub_download
|
| 5 |
import spaces
|
| 6 |
|
| 7 |
api = HfApi()
|
| 8 |
|
| 9 |
@spaces.GPU
|
| 10 |
-
def
|
| 11 |
-
if not token or not source_dataset or not target_dataset or not
|
| 12 |
return "❌ Error: Faltan datos obligatorios."
|
| 13 |
|
| 14 |
try:
|
| 15 |
-
|
| 16 |
-
file_path = file_path.strip().rstrip('.')
|
| 17 |
|
| 18 |
-
# 1.
|
| 19 |
-
|
| 20 |
-
repo_id=source_dataset,
|
| 21 |
-
filename=file_path,
|
| 22 |
-
repo_type="dataset",
|
| 23 |
-
token=token
|
| 24 |
-
)
|
| 25 |
|
| 26 |
-
#
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
"-c", "copy", "-movflags", "+faststart", output_path
|
| 31 |
]
|
| 32 |
-
subprocess.run(command, check=True)
|
| 33 |
|
| 34 |
-
#
|
| 35 |
-
|
| 36 |
-
path_or_fileobj=output_path,
|
| 37 |
-
path_in_repo=file_path,
|
| 38 |
-
repo_id=target_dataset,
|
| 39 |
-
repo_type="dataset",
|
| 40 |
-
token=token
|
| 41 |
-
)
|
| 42 |
|
| 43 |
-
|
| 44 |
-
|
|
|
|
|
|
|
|
|
|
| 45 |
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
except Exception as e:
|
| 49 |
-
return f"❌ Algo falló: {str(e)}"
|
| 50 |
|
| 51 |
-
# Interfaz gráfica adaptada para
|
| 52 |
with gr.Blocks() as demo:
|
| 53 |
-
gr.Markdown("# 🛠️ Migrador y Reparador
|
| 54 |
-
gr.Markdown("
|
| 55 |
|
| 56 |
token_input = gr.Textbox(label="Tu HF_TOKEN (Con permisos de WRITE)", type="password")
|
| 57 |
|
| 58 |
with gr.Row():
|
| 59 |
-
source_input = gr.Textbox(label="Dataset ORIGEN
|
| 60 |
-
target_input = gr.Textbox(label="Dataset DESTINO
|
| 61 |
|
| 62 |
-
|
|
|
|
|
|
|
|
|
|
| 63 |
|
| 64 |
-
btn = gr.Button("
|
| 65 |
-
output = gr.Textbox(label="Estado del proceso")
|
| 66 |
|
| 67 |
-
btn.click(fn=
|
| 68 |
|
| 69 |
demo.launch(theme=gr.themes.Soft())
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
import os
|
| 3 |
import subprocess
|
| 4 |
+
from huggingface_hub import HfApi, hf_hub_download, list_repo_files
|
| 5 |
import spaces
|
| 6 |
|
| 7 |
api = HfApi()
|
| 8 |
|
| 9 |
@spaces.GPU
|
| 10 |
+
def process_batch_videos(token, source_dataset, target_dataset, folder_path):
|
| 11 |
+
if not token or not source_dataset or not target_dataset or not folder_path:
|
| 12 |
return "❌ Error: Faltan datos obligatorios."
|
| 13 |
|
| 14 |
try:
|
| 15 |
+
folder_path = folder_path.strip().rstrip('/')
|
|
|
|
| 16 |
|
| 17 |
+
# 1. Listar todos los archivos del repo origen
|
| 18 |
+
all_files = list_repo_files(repo_id=source_dataset, repo_type="dataset", token=token)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
+
# Filtrar los archivos mp4 dentro de la carpeta especificada
|
| 21 |
+
mp4_files = [
|
| 22 |
+
f for f in all_files
|
| 23 |
+
if f.startswith(folder_path) and f.endswith('.mp4')
|
|
|
|
| 24 |
]
|
|
|
|
| 25 |
|
| 26 |
+
# Ordenar para consistencia y tomar estrictamente los primeros 3
|
| 27 |
+
mp4_files = sorted(mp4_files)[:3]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
|
| 29 |
+
if not mp4_files:
|
| 30 |
+
return f"⚠️ No se encontraron archivos .mp4 en la ruta: {folder_path}"
|
| 31 |
+
|
| 32 |
+
results = []
|
| 33 |
+
results.append(f"📁 Encontrados {len(mp4_files)} archivos para procesar (Modo WIP - Primeros 3):")
|
| 34 |
|
| 35 |
+
for idx, file_path in enumerate(mp4_files, 1):
|
| 36 |
+
results.append(f"\n--- Procesando [{idx}/3]: {file_path} ---")
|
| 37 |
+
|
| 38 |
+
# Descargar archivo original
|
| 39 |
+
local_path = hf_hub_download(
|
| 40 |
+
repo_id=source_dataset,
|
| 41 |
+
filename=file_path,
|
| 42 |
+
repo_type="dataset",
|
| 43 |
+
token=token
|
| 44 |
+
)
|
| 45 |
+
|
| 46 |
+
# Procesar con FFmpeg (Remux y Faststart)
|
| 47 |
+
output_path = f"video_arreglado_{idx}.mp4"
|
| 48 |
+
command = [
|
| 49 |
+
"ffmpeg", "-y", "-i", local_path,
|
| 50 |
+
"-c", "copy", "-movflags", "+faststart", output_path
|
| 51 |
+
]
|
| 52 |
+
subprocess.run(command, check=True)
|
| 53 |
+
|
| 54 |
+
# Subir al dataset destino limpio
|
| 55 |
+
api.upload_file(
|
| 56 |
+
path_or_fileobj=output_path,
|
| 57 |
+
path_in_repo=file_path,
|
| 58 |
+
repo_id=target_dataset,
|
| 59 |
+
repo_type="dataset",
|
| 60 |
+
token=token
|
| 61 |
+
)
|
| 62 |
+
|
| 63 |
+
# Limpieza local
|
| 64 |
+
if os.path.exists(output_path):
|
| 65 |
+
os.remove(output_path)
|
| 66 |
+
|
| 67 |
+
results.append(f"✅ ¡Reparado y subido con éxito a {target_dataset}!")
|
| 68 |
+
|
| 69 |
+
results.append("\n🎉 ¡Lote de prueba de los primeros 3 completado exitosamente!")
|
| 70 |
+
return "\n".join(results)
|
| 71 |
|
| 72 |
except Exception as e:
|
| 73 |
+
return f"❌ Algo falló en el proceso por lotes: {str(e)}"
|
| 74 |
|
| 75 |
+
# Interfaz gráfica adaptada para procesamiento por lotes (WIP)
|
| 76 |
with gr.Blocks() as demo:
|
| 77 |
+
gr.Markdown("# 🛠️ Migrador y Reparador Masivo (Modo WIP - Primeros 3)")
|
| 78 |
+
gr.Markdown("Escanea la carpeta, verifica y repara únicamente los primeros 3 archivos MP4 encontrados.")
|
| 79 |
|
| 80 |
token_input = gr.Textbox(label="Tu HF_TOKEN (Con permisos de WRITE)", type="password")
|
| 81 |
|
| 82 |
with gr.Row():
|
| 83 |
+
source_input = gr.Textbox(label="Dataset ORIGEN", value="xplhack1/bylatamsrc2")
|
| 84 |
+
target_input = gr.Textbox(label="Dataset DESTINO", value="orglatamsrc/latamsrcfixed")
|
| 85 |
|
| 86 |
+
folder_input = gr.Textbox(
|
| 87 |
+
label="Carpeta a escanear",
|
| 88 |
+
value="SERIES/EN_EMISION/Rivales_2024/TEMPORADA1"
|
| 89 |
+
)
|
| 90 |
|
| 91 |
+
btn = gr.Button("Verificar y Reparar Primeros 3 🚀", variant="primary")
|
| 92 |
+
output = gr.Textbox(label="Estado del proceso / Log", lines=15)
|
| 93 |
|
| 94 |
+
btn.click(fn=process_batch_videos, inputs=[token_input, source_input, target_input, folder_input], outputs=output)
|
| 95 |
|
| 96 |
demo.launch(theme=gr.themes.Soft())
|