Spaces:
Runtime error
Runtime error
Delete app (1).py
Browse files- app (1).py +0 -131
app (1).py
DELETED
@@ -1,131 +0,0 @@
|
|
1 |
-
import streamlit as st
|
2 |
-
import os
|
3 |
-
from pytube import YouTube
|
4 |
-
from bs4 import BeautifulSoup
|
5 |
-
from moviepy.editor import VideoFileClip
|
6 |
-
import whisper
|
7 |
-
import argostranslate.package
|
8 |
-
import argostranslate.translate
|
9 |
-
import torch
|
10 |
-
|
11 |
-
st.title("YouTube Video Translator")
|
12 |
-
form = st.form("input URL")
|
13 |
-
url = form.text_input('YouTube Video URL:',value='',placeholder='Paste your video URL here')
|
14 |
-
selected_language = st.selectbox("Select your desired language:", ["Arabic", "Chinese", "French", "German", "Hindi", "Italian", "Japanese", "Polish", "Russian", "Spanish"])
|
15 |
-
to_code = ''
|
16 |
-
if selected_language == "Arabic":
|
17 |
-
to_code = "ar"
|
18 |
-
elif selected_language == "Chinese":
|
19 |
-
to_code = "zh"
|
20 |
-
elif selected_language == "French":
|
21 |
-
to_code = "fr"
|
22 |
-
elif selected_language == "German":
|
23 |
-
to_code = "de"
|
24 |
-
elif selected_language == "Hindi":
|
25 |
-
to_code = "hi"
|
26 |
-
elif selected_language == "Italian":
|
27 |
-
to_code = "it"
|
28 |
-
elif selected_language == "Japanese":
|
29 |
-
to_code = "ja"
|
30 |
-
elif selected_language == "Polish":
|
31 |
-
to_code = "pl"
|
32 |
-
elif selected_language == "Russian":
|
33 |
-
to_code = "ru"
|
34 |
-
elif selected_language == "Spanish":
|
35 |
-
to_code = "es"
|
36 |
-
|
37 |
-
form.form_submit_button("Proceed")
|
38 |
-
|
39 |
-
|
40 |
-
def get_videotext(url, to_code):
|
41 |
-
print("to_Code: ", to_code)
|
42 |
-
download_dir_video = "downloaded_videos"
|
43 |
-
os.makedirs(download_dir_video, exist_ok=True)
|
44 |
-
download_dir_captions = "downloaded_captions"
|
45 |
-
os.makedirs(download_dir_captions, exist_ok=True)
|
46 |
-
yt = YouTube(url)
|
47 |
-
video = yt.streams.filter(file_extension='mp4').first()
|
48 |
-
print("yt: ", yt.streams.first())
|
49 |
-
video.download(output_path=download_dir_video)
|
50 |
-
#get audio
|
51 |
-
audio_dir = "audio"
|
52 |
-
os.makedirs(audio_dir, exist_ok=True)
|
53 |
-
|
54 |
-
video_filename = os.listdir(download_dir_video)[0]
|
55 |
-
audio_file_name = os.path.splitext(video_filename)[0]
|
56 |
-
video_file_path = os.path.join(download_dir_video, video_filename)
|
57 |
-
audio_file_path = os.path.join(audio_dir, f"{audio_file_name}.wav")
|
58 |
-
|
59 |
-
video = VideoFileClip(video_file_path)
|
60 |
-
video.audio.write_audiofile(audio_file_path)
|
61 |
-
|
62 |
-
##audio to text
|
63 |
-
model = whisper.load_model("base")
|
64 |
-
|
65 |
-
result = model.transcribe(os.path.join(audio_dir, f"{audio_file_name}.wav"))
|
66 |
-
segments = result.get("segments", [])
|
67 |
-
|
68 |
-
time_results = []
|
69 |
-
text_results = []
|
70 |
-
|
71 |
-
for segment in segments:
|
72 |
-
start_time_sec = segment.get("start", 0.0)
|
73 |
-
start_time_formatted = "{:02d}:{:02d}:{:06.3f}".format(
|
74 |
-
int(start_time_sec // 3600),
|
75 |
-
int((start_time_sec % 3600) // 60),
|
76 |
-
start_time_sec % 60
|
77 |
-
)
|
78 |
-
time_results.append(start_time_formatted)
|
79 |
-
text_results.append(segment.get("text", "").strip())
|
80 |
-
|
81 |
-
from_code = "en"
|
82 |
-
# to_code = "fr"
|
83 |
-
|
84 |
-
argostranslate.package.update_package_index()
|
85 |
-
available_packages = argostranslate.package.get_available_packages()
|
86 |
-
package_to_install = next(
|
87 |
-
filter(
|
88 |
-
lambda x: x.from_code == from_code and x.to_code == to_code, available_packages
|
89 |
-
)
|
90 |
-
)
|
91 |
-
argostranslate.package.install_from_path(package_to_install.download())
|
92 |
-
|
93 |
-
translated_text_results = [argostranslate.translate.translate(text, from_code, to_code) for text in text_results]
|
94 |
-
|
95 |
-
|
96 |
-
return time_results, text_results, translated_text_results
|
97 |
-
|
98 |
-
print(url)
|
99 |
-
column_width = 100
|
100 |
-
with st.spinner(f'Please wait while we create translations in {selected_language}...'):
|
101 |
-
if url:
|
102 |
-
time_results, text_results, translated_text_results= get_videotext(url, to_code)
|
103 |
-
|
104 |
-
table_style = """
|
105 |
-
<style>
|
106 |
-
table {
|
107 |
-
width: 100%;
|
108 |
-
border-collapse: collapse;
|
109 |
-
}
|
110 |
-
th, td {
|
111 |
-
padding: 8px;
|
112 |
-
text-align: left;
|
113 |
-
}
|
114 |
-
th {
|
115 |
-
background-color: #f2f2f2;
|
116 |
-
color: black;
|
117 |
-
}
|
118 |
-
</style>
|
119 |
-
"""
|
120 |
-
|
121 |
-
|
122 |
-
table_header = "<tr><th>TimeStamp</th><th>Original</th><th>Translated</th></tr>"
|
123 |
-
|
124 |
-
table_rows = "".join(f"<tr><td>{time}</td><td>{text}</td><td>{translated_text}</td></tr>"
|
125 |
-
for time, text, translated_text in zip(time_results, text_results, translated_text_results))
|
126 |
-
|
127 |
-
table_html = f"<table>{table_header}{table_rows}</table>"
|
128 |
-
|
129 |
-
|
130 |
-
st.write(table_style, unsafe_allow_html=True)
|
131 |
-
st.write(table_html, unsafe_allow_html=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|