-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathapp.py
191 lines (162 loc) · 9.67 KB
/
app.py
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
import streamlit as st
from dotenv import load_dotenv
load_dotenv()
import os
import google.generativeai as genai
from youtube_transcript_api import YouTubeTranscriptApi
import requests
from io import BytesIO
from PIL import Image
import json
import base64
import pandas as pd
import matplotlib.pyplot as plt
genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
def extract_video_id(youtube_url):
if 'youtu.be' in youtube_url:
return youtube_url.split('/')[-1]
elif 'v=' in youtube_url:
return youtube_url.split('v=')[1].split('&')[0]
else:
return None
def extract_transcript(video_id, language='en'):
try:
transcript_text = YouTubeTranscriptApi.get_transcript(video_id, languages=[language])
return transcript_text
except Exception as e:
st.error(f"Failed to fetch transcript: {str(e)}")
return None
def generate_content(transcript, prompt, tone, length, model_name="gemini-1.5-pro-latest"):
model = genai.GenerativeModel(model_name)
transcript_text = " ".join([i["text"] for i in transcript])
full_prompt = f"{prompt}\n\nTranscript:\n{transcript_text}\n\nTone: {tone}\nLength: {length}"
response = model.generate_content(full_prompt)
return response.text
def get_video_metadata(video_id):
url = f"https://www.googleapis.com/youtube/v3/videos?part=snippet&id={video_id}&key={os.getenv('YOUTUBE_API_KEY')}"
response = requests.get(url)
if response.status_code == 200:
data = response.json()
if data["items"]:
return data["items"][0]["snippet"]
return None
def generate_srt(transcript):
srt = ""
for i, segment in enumerate(transcript, 1):
start = segment['start']
duration = segment['duration']
end = start + duration
text = segment['text']
srt += f"{i}\n{start:.3f} --> {end:.3f}\n{text}\n\n"
return srt
def get_table_download_link(df):
csv = df.to_csv(index=False)
b64 = base64.b64encode(csv.encode()).decode()
href = f'<a href="data:file/csv;base64,{b64}" download="transcript_analysis.csv">Download Transcript Analysis (CSV)</a>'
return href
def main():
st.set_page_config(page_title="Advanced Global Content Studio", page_icon="🌐", layout="wide")
st.title("🌐 YouTube to Global Multimedia Studio")
st.markdown("Transform YouTube videos into rich, multilingual content with advanced analytics! 🚀🌍")
youtube_link = st.text_input("Enter YouTube Video Link:", placeholder="https://www.youtube.com/watch?v=dQw4w9WgXcQ")
if youtube_link:
video_id = extract_video_id(youtube_link)
if video_id:
metadata = get_video_metadata(video_id)
if metadata:
st.header(metadata["title"])
st.image(metadata["thumbnails"]["high"]["url"], use_column_width=True)
st.write(f"**Channel:** {metadata['channelTitle']} | **Published:** {metadata['publishedAt'][:10]}")
st.write(metadata["description"])
col1, col2, col3 = st.columns(3)
with col1:
content_type = st.selectbox("Select Content Type:", ["News Article", "Blog Post", "Social Media Post", "Email Newsletter", "Product Description", "Academic Summary"])
with col2:
tone = st.selectbox("Select Tone:", ["Informative", "Casual", "Professional", "Humorous", "Inspirational", "Empathetic", "Persuasive"])
with col3:
languages = {
"English": "en", "Spanish": "es", "French": "fr", "German": "de", "Italian": "it",
"Japanese": "ja", "Korean": "ko", "Portuguese": "pt", "Russian": "ru",
"Chinese (Simplified)": "zh-Hans", "Chinese (Traditional)": "zh-Hant",
"Arabic": "ar", "Hindi": "hi", "Bengali": "bn", "Urdu": "ur", "Turkish": "tr",
"Dutch": "nl", "Polish": "pl", "Swedish": "sv", "Thai": "th", "Vietnamese": "vi",
"Greek": "el", "Hebrew": "he", "Indonesian": "id", "Malay": "ms", "Filipino": "fil"
}
input_language = st.selectbox("Select Transcript Language:", list(languages.keys()))
length = st.slider("Content Length:", min_value=100, max_value=5000, value=500, step=100)
col4, col5 = st.columns(2)
with col4:
model_option = st.radio("Choose Model:", ["gemini-1.5-pro-latest", "gemini-1.5-flash-latest", "gemini-1.0-pro"])
with col5:
analyze_transcript = st.checkbox("Analyze Transcript")
output_language = st.selectbox("Select Output Language:", list(languages.keys()))
if st.button("Generate Content 🖋️"):
with st.spinner(f"Fetching video transcript in {input_language}..."):
transcript = extract_transcript(video_id, languages[input_language])
if transcript:
# Download SRT
srt_data = generate_srt(transcript)
st.download_button(
label="Download SRT 📝",
data=srt_data,
file_name=f"transcript_{languages[input_language]}.srt",
mime="text/plain"
)
if analyze_transcript:
with st.expander("Transcript Analysis 📊"):
analysis_prompt = f"Analyze the following transcript for key themes, sentiment, and important timestamps. Transcript: {json.dumps(transcript)}"
analysis = generate_content(transcript, analysis_prompt, "Informative", 1000, model_option)
st.write(analysis)
# Word frequency analysis
from collections import Counter
words = " ".join([seg["text"] for seg in transcript]).lower().split()
word_freq = Counter(words)
df = pd.DataFrame.from_dict(word_freq, orient='index', columns=['Frequency']).sort_values(by='Frequency', ascending=False)
plt.figure(figsize=(10, 6))
df.head(20).plot(kind='bar')
plt.title("Top 20 Word Frequencies")
plt.xlabel("Words")
plt.ylabel("Frequency")
st.pyplot(plt)
st.markdown(get_table_download_link(df), unsafe_allow_html=True)
with st.spinner(f"Generating {content_type.lower()} in {output_language}..."):
prompts = {
"News Article": f"You are a news writer. Generate a concise, factual news article in a journalistic style based on the given YouTube video transcript. Write the article in {output_language}.",
"Blog Post": f"You are a blog writer. Create a detailed, engaging blog post covering the key points and insights from the given YouTube video transcript. Write the blog post in {output_language}.",
"Social Media Post": f"You are a social media manager. Create a short, engaging post suitable for platforms like Twitter or LinkedIn, highlighting the main takeaway from the YouTube video. Write the post in {output_language}.",
"Email Newsletter": f"You are a content marketer. Create an informative and engaging email newsletter summarizing the key points from the YouTube video. Write the newsletter in {output_language}.",
"Product Description": f"You are a copywriter. Based on the YouTube video, create a compelling product description that highlights features and benefits. Write in {output_language}.",
"Academic Summary": f"You are an academic researcher. Provide a concise, scholarly summary of the key arguments and evidence presented in the YouTube video. Write in {output_language}."
}
content = generate_content(transcript, prompts[content_type], tone, length, model_option)
st.subheader(f"Generated {content_type} in {output_language} 📝")
st.write(content)
st.download_button(
label=f"Download {content_type} 📥",
data=content,
file_name=f"{content_type.replace(' ', '_').lower()}_{output_language}.txt",
mime="text/plain"
)
else:
st.warning(f"Please make sure the video has captions available in {input_language}.")
else:
st.error("Invalid YouTube URL. Please enter a valid video link.")
st.sidebar.title("About")
st.sidebar.info(
"This advanced global studio uses AI to transform YouTube videos into rich, multilingual content with analytics. "
"Perfect for international content creators, global marketers, researchers, and anyone looking to deeply understand and reach a diverse audience! 🌐🎓🚀\n\n"
"Made with ❤️ by Ragib"
)
st.sidebar.title("New Features")
st.sidebar.markdown(
"- 🚀 Fast `gemini-1.5-flash-latest` model\n"
"- 📊 Transcript analysis with word frequency\n"
"- 📝 SRT subtitle file download\n"
"- 🎓 Academic summary generation\n"
"- 🛍️ Product descriptions from videos\n"
"- 🌐 Support for Greek, Hebrew, Indonesian & more\n"
"- 📈 Longer content up to 5000 characters\n"
"- 🎨 'Persuasive' tone for marketing content"
)
if __name__ == "__main__":
main()