File size: 11,171 Bytes
c94aadc
 
 
 
 
306cc5c
1caaaa3
c94aadc
 
 
 
 
 
 
306cc5c
319260b
 
 
c94aadc
abed2e9
c94aadc
 
 
 
 
6065fb3
c94aadc
f24aa65
c94aadc
f24aa65
c94aadc
6065fb3
 
bdb3cfa
f24aa65
2acde19
f24aa65
 
 
 
 
 
 
 
6065fb3
306cc5c
4cf7ad9
 
 
 
 
 
 
 
84bda60
 
 
2acde19
 
4cf7ad9
 
 
 
2acde19
4cf7ad9
 
 
 
 
84bda60
4cf7ad9
 
6065fb3
4cf7ad9
9ea2009
f24aa65
2acde19
4cf7ad9
 
 
 
 
 
 
 
 
 
 
f24aa65
c94aadc
2acde19
4cf7ad9
 
2acde19
4cf7ad9
 
 
2acde19
4cf7ad9
 
 
 
 
 
 
 
 
6065fb3
4cf7ad9
2acde19
4cf7ad9
 
 
 
2acde19
4cf7ad9
 
 
9ea2009
c94aadc
2acde19
4cf7ad9
 
 
 
 
 
 
 
 
 
 
9ea2009
c94aadc
2acde19
c94aadc
 
2acde19
abed2e9
2acde19
 
c94aadc
abed2e9
2acde19
c94aadc
 
 
 
abed2e9
2acde19
78a3857
 
2acde19
 
78a3857
2acde19
 
78a3857
 
 
 
 
2acde19
78a3857
 
 
 
9ea2009
319260b
2acde19
319260b
 
 
 
 
2acde19
319260b
 
 
 
 
 
2acde19
319260b
 
 
 
 
2acde19
319260b
 
 
 
 
 
 
c94aadc
2acde19
c94aadc
6065fb3
c94aadc
 
 
 
 
 
6065fb3
c94aadc
 
 
6065fb3
c94aadc
 
 
6065fb3
c94aadc
 
 
6065fb3
c94aadc
 
 
6065fb3
78a3857
 
 
 
 
 
 
 
 
6065fb3
0a28283
6065fb3
c94aadc
 
 
 
 
 
 
 
59b95fb
 
c94aadc
 
 
 
 
 
6065fb3
 
c94aadc
 
 
 
319260b
c94aadc
 
6065fb3
c94aadc
6065fb3
 
78a3857
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
import gradio as gr
from tts_module import get_voices, text_to_speech
from pixabay_api import search_pixabay
from moviepy.editor import (
    AudioFileClip, VideoFileClip, CompositeAudioClip,
    concatenate_audioclips, concatenate_videoclips, vfx, CompositeVideoClip,
    ColorClip
)
import asyncio
import os
import json
import time
import requests
import random
from PIL import Image  # Para manejar imágenes correctamente
from googleapiclient.discovery import build
from google.oauth2 import service_account
from googleapiclient.http import MediaFileUpload

# Crear el archivo de credenciales de servicio desde los secretos
service_account_info = json.loads(os.getenv('GOOGLE_SERVICE_ACCOUNT', '{}'))
if service_account_info:
    with open('service-account.json', 'w') as f:
        json.dump(service_account_info, f)

# Define la carpeta de salida y la carpeta temporal
output_folder = "outputs"
temp_dir = "temp_files"
os.makedirs(output_folder, exist_ok=True)
os.makedirs(temp_dir, exist_ok=True)

# ID de la carpeta de destino en Google Drive
FOLDER_ID = "12S6adpanAXjf71pKKGRRPqpzbJa5XEh3"  # Reemplaza con tu ID de carpeta

def cleanup_temp_files():
    """Elimina todos los archivos temporales de la carpeta temp_files."""
    for filename in os.listdir(temp_dir):
        file_path = os.path.join(temp_dir, filename)
        try:
            if os.path.isfile(file_path):
                os.remove(file_path)
        except Exception as e:
            print(f"Error deleting {file_path}: {e}")

def resize_video(clip, target_width=1920, target_height=1080):
    """Redimensiona el video al tamaño 1080p (16:9)."""
    try:
        w, h = clip.size
        current_aspect_ratio = w / h
        target_aspect_ratio = target_width / target_height

        if abs(current_aspect_ratio - target_aspect_ratio) < 0.1:
            return clip.resize((target_width, target_height))

        # Crear un fondo negro con las dimensiones objetivo
        background = ColorClip(size=(target_width, target_height), color=[0, 0, 0]).set_duration(clip.duration)

        # Redimensionar el video original para mantener su proporción
        if current_aspect_ratio < target_aspect_ratio:  # Video vertical
            new_height = target_height
            new_width = int(new_height * current_aspect_ratio)
            x_center = (target_width - new_width) / 2
            resized_clip = clip.resize(width=new_width).set_position((x_center, 0))
        else:  # Video horizontal
            new_width = target_width
            new_height = int(new_width / current_aspect_ratio)
            y_center = (target_height - new_height) / 2
            resized_clip = clip.resize(height=new_height).set_position((0, y_center))

        # Combinar el fondo negro con el video redimensionado
        return CompositeVideoClip([background, resized_clip], size=(target_width, target_height))
    except Exception as e:
        print(f"Error en resize_video: {e}")
        return clip

def download_video(link):
    """Descarga un video desde un enlace y lo guarda en la carpeta temporal."""
    try:
        video_response = requests.get(link)
        if video_response.status_code != 200:
            return None
        temp_video_path = os.path.join(temp_dir, f"temp_video_{int(time.time())}.mp4")
        with open(temp_video_path, "wb") as f:
            f.write(video_response.content)
        return temp_video_path
    except Exception as e:
        print(f"Error downloading video: {e}")
        return None

def concatenate_pixabay_videos(keywords, num_videos_per_keyword=1):
    """Concatena videos de Pixabay basados en palabras clave."""
    keyword_list = [keyword.strip() for keyword in keywords.split(",") if keyword.strip()]
    if not keyword_list:
        keyword_list = ["nature"]  # Palabra clave por defecto
    video_clips = []
    for keyword in keyword_list:
        try:
            print(f"Buscando videos para la palabra clave '{keyword}'...")
            links = search_pixabay(keyword, num_results=num_videos_per_keyword)
            if not links:
                print(f"No se encontraron videos para '{keyword}', probando con 'nature'")
                links = search_pixabay("nature", num_results=num_videos_per_keyword)
                if not links:
                    continue
            temp_video_path = download_video(links[0])
            if temp_video_path:
                clip = VideoFileClip(temp_video_path)
                processed_clip = resize_video(clip)
                video_clips.append(processed_clip)
                os.remove(temp_video_path)  # Eliminar archivo temporal después de usarlo
        except Exception as e:
            print(f"Error procesando palabra clave '{keyword}': {e}")
            continue
    if not video_clips:
        # Si no hay videos, creamos un clip negro de 5 segundos
        return ColorClip(size=(1920, 1080), color=[0, 0, 0], duration=5)
    random.shuffle(video_clips)
    return concatenate_videoclips(video_clips, method="compose")

def adjust_background_music(video_duration, music_file):
    """Ajusta la música de fondo para que coincida con la duración del video."""
    try:
        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)
        music = music.subclip(0, video_duration)
        return music.volumex(0.2)
    except Exception as e:
        print(f"Error ajustando música: {e}")
        return None

def combine_audio_video(audio_file, video_clip, music_clip=None):
    """Combina el audio y el video en un archivo final."""
    try:
        audio_clip = AudioFileClip(audio_file)
        total_duration = audio_clip.duration + 2  # Añadimos 2 segundos extra

        # Aseguramos que el video tenga la duración correcta
        video_clip = video_clip.loop(duration=total_duration)
        video_clip = video_clip.set_duration(total_duration).fadeout(2)

        # Combinamos el audio principal
        final_clip = video_clip.set_audio(audio_clip)
        if music_clip:
            music_clip = music_clip.set_duration(total_duration).audio_fadeout(2)
            final_clip = final_clip.set_audio(CompositeAudioClip([audio_clip, music_clip]))

        # Generamos el nombre del archivo y la ruta
        output_filename = f"final_video_{int(time.time())}.mp4"
        output_path = os.path.join(output_folder, output_filename)

        # Guardamos el video
        final_clip.write_videofile(output_path, codec="libx264", audio_codec="aac", fps=24)

        # Limpiamos los clips
        final_clip.close()
        video_clip.close()
        audio_clip.close()
        if music_clip:
            music_clip.close()

        return output_path
    except Exception as e:
        print(f"Error combinando audio y video: {e}")
        return None

def upload_to_google_drive(file_path, folder_id):
    """Sube un archivo a Google Drive y devuelve el enlace público."""
    try:
        creds = service_account.Credentials.from_service_account_file(
            'service-account.json', scopes=['https://www.googleapis.com/auth/drive']
        )
        service = build('drive', 'v3', credentials=creds)

        file_metadata = {
            'name': os.path.basename(file_path),
            'parents': [folder_id]
        }
        media = MediaFileUpload(file_path, resumable=True)
        file = service.files().create(body=file_metadata, media_body=media, fields='id').execute()

        permission = {
            'type': 'anyone',
            'role': 'reader'
        }
        service.permissions().create(fileId=file['id'], body=permission).execute()

        file_id = file['id']
        download_link = f"https://drive.google.com/uc?export=download&id={file_id}"
        return download_link
    except Exception as e:
        print(f"Error subiendo a Google Drive: {e}")
        return None

def process_input(text, txt_file, mp3_file, selected_voice, rate, pitch, keywords):
    """Procesa la entrada del usuario y genera el video final."""
    try:
        # Determinamos el texto a usar
        if text.strip():
            final_text = text
        elif txt_file is not None:
            final_text = txt_file.decode("utf-8")
        else:
            raise ValueError("No text input provided")
        # Generamos el audio
        audio_file = asyncio.run(text_to_speech(final_text, selected_voice, rate, pitch))
        if not audio_file:
            raise ValueError("Failed to generate audio")
        # Generamos el video
        video_clip = concatenate_pixabay_videos(keywords, num_videos_per_keyword=1)
        if not video_clip:
            raise ValueError("Failed to generate video")
        # Procesamos la música de fondo si existe
        music_clip = None
        if mp3_file is not None:
            music_clip = adjust_background_music(video_clip.duration, mp3_file.name)
        # Combinamos todo
        final_video_path = combine_audio_video(audio_file, video_clip, music_clip)
        if not final_video_path:
            raise ValueError("Failed to combine audio and video")
        # Subimos a Google Drive y obtenemos el enlace
        download_link = upload_to_google_drive(final_video_path, folder_id=FOLDER_ID)
        if download_link:
            return f"[Descargar video]({download_link})"
        else:
            raise ValueError("Error subiendo el video a Google Drive")
    except Exception as e:
        print(f"Error durante el procesamiento: {e}")
        return None
    finally:
        cleanup_temp_files()  # Limpiar archivos temporales al finalizar

# Interfaz Gradio
with gr.Blocks() as demo:
    gr.Markdown("# Text-to-Video Generator")
    with gr.Row():
        with gr.Column():
            text_input = gr.Textbox(label="Write your text here", lines=5)
            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"])
            keyword_input = gr.Textbox(
                label="Enter keywords separated by commas",
                value="fear,religion,god,demons,aliens,possession"
            )
            voices = asyncio.run(get_voices())
            voice_dropdown = gr.Dropdown(choices=list(voices.keys()), label="Select Voice")
            rate_slider = gr.Slider(minimum=-50, maximum=50, value=0, label="Speech Rate Adjustment (%)", step=1)
            pitch_slider = gr.Slider(minimum=-20, maximum=20, value=0, label="Pitch Adjustment (Hz)", step=1)
        with gr.Column():
            output_link = gr.Markdown("")  # Mostrar el enlace de descarga

    btn = gr.Button("Generate Video")
    btn.click(
        process_input,
        inputs=[text_input, txt_file_input, mp3_file_input, voice_dropdown, rate_slider, pitch_slider, keyword_input],
        outputs=output_link
    )

# Leer el puerto asignado por Hugging Face
port = int(os.getenv("PORT", 7860))

# Lanzar la aplicación
demo.launch(server_name="0.0.0.0", server_port=port, share=True, show_error=True)