|
|
|
|
|
|
|
|
|
|
|
|
|
import pathlib
|
|
import sys, os
|
|
from gtts import gTTS
|
|
import gradio as gr
|
|
import os
|
|
import speech_recognition as sr
|
|
from googletrans import Translator, constants
|
|
from pprint import pprint
|
|
from moviepy.editor import *
|
|
from pytube import YouTube
|
|
from youtube_transcript_api import YouTubeTranscriptApi
|
|
from utils import *
|
|
|
|
def download_video(url):
|
|
print("Downloading...")
|
|
local_file = (
|
|
YouTube(url)
|
|
.streams.filter(progressive=True, file_extension="mp4")
|
|
.first()
|
|
.download()
|
|
)
|
|
print("Downloaded")
|
|
return local_file
|
|
|
|
def validate_url(url):
|
|
import validators
|
|
if not validators.url(url):
|
|
print("Hi there URL seems invalid ")
|
|
|
|
|
|
def cleanup():
|
|
import pathlib
|
|
import glob
|
|
types = ('*.mp4', '*.wav')
|
|
|
|
junks = []
|
|
for files in types:
|
|
junks.extend(glob.glob(files))
|
|
try:
|
|
|
|
for junk in junks:
|
|
print("Deleting",junk)
|
|
|
|
file = pathlib.Path(junk)
|
|
|
|
file.unlink()
|
|
except Exception:
|
|
print("I cannot delete the file because it is being used by another process")
|
|
|
|
def getSize(filename):
|
|
st = os.stat(filename)
|
|
return st.st_size
|
|
|
|
|
|
def generate_transcript(url,lang_api):
|
|
id = url[url.index("=")+1:]
|
|
transcript = YouTubeTranscriptApi.get_transcript(id,languages=[lang_api])
|
|
script = ""
|
|
for text in transcript:
|
|
t = text["text"]
|
|
if t != '[Music]':
|
|
script += t + " "
|
|
return script
|
|
|
|
|
|
def video_to_translate(url,initial_language,final_language):
|
|
|
|
|
|
if initial_language == "English":
|
|
lang_in='en-US'
|
|
lang_api='en'
|
|
elif initial_language == "Italian":
|
|
lang_in='it-IT'
|
|
lang_api='it'
|
|
elif initial_language == "Spanish":
|
|
lang_in='es-MX'
|
|
lang_api='es'
|
|
elif initial_language == "Russian":
|
|
lang_in='ru-RU'
|
|
lang_api='rus'
|
|
elif initial_language == "German":
|
|
lang_in='de-DE'
|
|
lang_api='de'
|
|
elif initial_language == "Japanese":
|
|
lang_in='ja-JP'
|
|
lang_api='ja'
|
|
if final_language == "English":
|
|
lang='en'
|
|
elif final_language == "Italian":
|
|
lang='it'
|
|
elif final_language == "Spanish":
|
|
lang='es'
|
|
elif final_language == "Russian":
|
|
lang='ru'
|
|
elif final_language == "German":
|
|
lang='de'
|
|
elif final_language == "Japanese":
|
|
lang='ja'
|
|
|
|
|
|
home_dir = os.getcwd()
|
|
print('Initial directory:',home_dir)
|
|
cleanup()
|
|
|
|
temp_dir=os.path.join(home_dir, "temp")
|
|
print('Temporal directory:',temp_dir)
|
|
|
|
pathlib.Path(temp_dir).mkdir(parents=True, exist_ok=True)
|
|
|
|
os.chdir(temp_dir)
|
|
print('Changing temporal directory',os.getcwd())
|
|
|
|
cleanup()
|
|
file_obj=download_video(url)
|
|
print(file_obj)
|
|
|
|
videoclip = VideoFileClip(file_obj)
|
|
try:
|
|
|
|
text = generate_transcript(url,lang_api)
|
|
print("Transcript Found")
|
|
except Exception:
|
|
print("No Transcript Found")
|
|
|
|
|
|
videoclip.audio.write_audiofile("audio.wav",codec='pcm_s16le')
|
|
|
|
r = sr.Recognizer()
|
|
|
|
with sr.AudioFile("audio.wav") as source:
|
|
|
|
audio_data = r.record(source)
|
|
|
|
print("Recognize from ",lang_in)
|
|
|
|
size_wav=getSize("audio.wav")
|
|
if size_wav > 50000000:
|
|
print("The wav is too large")
|
|
audio_chunks=split_audio_wav("audio.wav")
|
|
text=""
|
|
for chunk in audio_chunks:
|
|
print("Converting audio to text",chunk)
|
|
try:
|
|
text_chunk= r.recognize_google(audio_data, language = lang_in)
|
|
except Exception:
|
|
print("This video cannot be recognized")
|
|
cleanup()
|
|
|
|
os.chdir(home_dir)
|
|
return "./demo/tryagain.mp4"
|
|
text=text+text_chunk+" "
|
|
text=str(text)
|
|
print(type(text))
|
|
|
|
else:
|
|
text = r.recognize_google(audio_data, language = lang_in)
|
|
|
|
print("Destination language ",lang)
|
|
|
|
|
|
translator = Translator()
|
|
|
|
|
|
try:
|
|
translation = translator.translate(text, dest=lang)
|
|
except Exception:
|
|
print("This text cannot be translated")
|
|
cleanup()
|
|
|
|
os.chdir(home_dir)
|
|
return "./demo/tryagain.mp4"
|
|
|
|
|
|
trans=translation.text
|
|
|
|
myobj = gTTS(text=trans, lang=lang, slow=False)
|
|
myobj.save("audio.wav")
|
|
|
|
audioclip = AudioFileClip("audio.wav")
|
|
|
|
|
|
new_audioclip = CompositeAudioClip([audioclip])
|
|
videoclip.audio = new_audioclip
|
|
new_video="video_translated_"+lang+".mp4"
|
|
|
|
|
|
os.chdir(home_dir)
|
|
print('Final directory',os.getcwd())
|
|
|
|
videoclip.write_videofile(new_video)
|
|
|
|
videoclip.close()
|
|
del file_obj
|
|
|
|
return new_video
|
|
|
|
initial_language = gr.inputs.Dropdown(["English","Italian","Japanese","Russian","Spanish","German"])
|
|
final_language = gr.inputs.Dropdown([ "Russian","Italian","Spanish","German","English","Japanese"])
|
|
url =gr.inputs.Textbox(label = "Enter the YouTube URL below:")
|
|
|
|
|
|
gr.Interface(fn = video_to_translate,
|
|
inputs = [url,initial_language,final_language],
|
|
outputs = 'video',
|
|
verbose = True,
|
|
title = 'Video Youtube Translator',
|
|
description = 'A simple application that translates Youtube videos from English, Italian, Japanese, Russian, Spanish, and German to Italian, Spanish, Russian, English and Japanese. Wait one minute to process.',
|
|
article =
|
|
'''<div>
|
|
<p style="text-align: center"> All you need to do is to paste the Youtube link and hit submit, then wait for compiling. After that click on Play/Pause for listing to the video. The video is saved in an mp4 format.
|
|
For more information visit <a href="https://ruslanmv.com/">ruslanmv.com</a>
|
|
</p>
|
|
</div>''',
|
|
|
|
examples = [
|
|
["https://www.youtube.com/watch?v=Cu3R5it4cQs&list", "English","Italian"],
|
|
["https://www.youtube.com/watch?v=fkGCLIQx1MI", "English","Spanish"],
|
|
["https://www.youtube.com/watch?v=fkGCLIQx1MI", "English","Russian"],
|
|
["https://www.youtube.com/watch?v=_5YeX8eCLgA&ab_channel=TheTelegraph", "Russian","English"],
|
|
["https://www.youtube.com/watch?v=qzzweIQoIOU", "Japanese","English"],
|
|
["https://www.youtube.com/watch?v=eo17uDr2_XA", "German","Spanish"]
|
|
]
|
|
).launch() |
