Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -10,25 +10,36 @@ import json
|
|
10 |
from datetime import datetime
|
11 |
import google.generativeai as genai
|
12 |
|
13 |
-
# Configure Gemini
|
14 |
-
genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
|
15 |
-
model = genai.GenerativeModel(model_name="gemini-1.5-flash-latest")
|
16 |
-
|
17 |
class YouTubeDownloader:
|
18 |
def __init__(self):
|
19 |
self.download_dir = tempfile.mkdtemp()
|
|
|
20 |
self.temp_downloads = tempfile.mkdtemp(prefix="youtube_downloads_")
|
|
|
21 |
self.downloads_folder = os.path.join(os.path.expanduser("~"), "Downloads", "YouTube_Downloads")
|
22 |
os.makedirs(self.downloads_folder, exist_ok=True)
|
23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
24 |
def cleanup(self):
|
|
|
25 |
try:
|
26 |
if hasattr(self, 'download_dir') and os.path.exists(self.download_dir):
|
27 |
shutil.rmtree(self.download_dir)
|
|
|
28 |
if hasattr(self, 'temp_downloads') and os.path.exists(self.temp_downloads):
|
29 |
shutil.rmtree(self.temp_downloads)
|
|
|
30 |
except Exception as e:
|
31 |
-
print(f"β οΈ
|
32 |
|
33 |
def is_valid_youtube_url(self, url):
|
34 |
youtube_regex = re.compile(
|
@@ -38,50 +49,166 @@ class YouTubeDownloader:
|
|
38 |
return youtube_regex.match(url) is not None
|
39 |
|
40 |
def generate_scene_breakdown_gemini(self, video_info):
|
41 |
-
|
42 |
-
|
43 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
44 |
|
|
|
|
|
|
|
|
|
|
|
45 |
if not duration:
|
46 |
-
return ["[Duration Unknown]
|
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 |
-
def
|
73 |
-
|
74 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
75 |
|
76 |
-
|
77 |
-
|
78 |
-
|
79 |
-
|
80 |
-
return
|
81 |
-
|
82 |
-
return
|
|
|
|
|
|
|
83 |
|
84 |
def format_video_info(self, video_info):
|
|
|
|
|
|
|
|
|
|
|
85 |
title = video_info.get("title", "Unknown")
|
86 |
uploader = video_info.get("uploader", "Unknown")
|
87 |
duration = video_info.get("duration", 0)
|
@@ -90,48 +217,386 @@ Now generate the scene breakdown:
|
|
90 |
like_count = video_info.get("like_count", 0)
|
91 |
comment_count = video_info.get("comment_count", 0)
|
92 |
upload_date = video_info.get("upload_date", "Unknown")
|
93 |
-
|
|
|
|
|
|
|
|
|
|
|
94 |
|
|
|
95 |
scene_descriptions = self.generate_scene_breakdown_gemini(video_info)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96 |
|
97 |
-
def fmt(n):
|
98 |
-
if not n: return "0"
|
99 |
-
if n > 1_000_000: return f"{n/1_000_000:.1f}M"
|
100 |
-
if n > 1_000: return f"{n/1_000:.1f}K"
|
101 |
-
return str(n)
|
102 |
-
|
103 |
-
summary = f"""
|
104 |
π BASIC INFORMATION
|
105 |
-
{'β'*
|
106 |
-
πΉ Title
|
107 |
-
π€ Uploader
|
108 |
-
π
Upload Date
|
109 |
-
β±οΈ Duration
|
110 |
-
|
111 |
-
|
112 |
-
|
113 |
-
|
114 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
115 |
{'β'*30}
|
116 |
{chr(10).join(scene_descriptions)}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
117 |
"""
|
118 |
-
|
|
|
119 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
120 |
downloader = YouTubeDownloader()
|
121 |
|
122 |
-
def
|
123 |
-
|
124 |
-
if not
|
125 |
-
return
|
126 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
127 |
|
128 |
-
|
129 |
-
|
130 |
-
|
131 |
-
|
132 |
-
|
133 |
-
|
134 |
-
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
135 |
|
136 |
if __name__ == "__main__":
|
137 |
-
|
|
|
|
|
|
|
|
10 |
from datetime import datetime
|
11 |
import google.generativeai as genai
|
12 |
|
|
|
|
|
|
|
|
|
13 |
class YouTubeDownloader:
|
14 |
def __init__(self):
|
15 |
self.download_dir = tempfile.mkdtemp()
|
16 |
+
# Use temp directory for Gradio compatibility
|
17 |
self.temp_downloads = tempfile.mkdtemp(prefix="youtube_downloads_")
|
18 |
+
# Also create user downloads folder for copying
|
19 |
self.downloads_folder = os.path.join(os.path.expanduser("~"), "Downloads", "YouTube_Downloads")
|
20 |
os.makedirs(self.downloads_folder, exist_ok=True)
|
21 |
+
self.gemini_model = None
|
22 |
+
|
23 |
+
def configure_gemini(self, api_key):
|
24 |
+
"""Configure Gemini API with the provided key"""
|
25 |
+
try:
|
26 |
+
genai.configure(api_key=api_key)
|
27 |
+
self.gemini_model = genai.GenerativeModel(model_name="gemini-1.5-flash-latest")
|
28 |
+
return True, "β
Gemini API configured successfully!"
|
29 |
+
except Exception as e:
|
30 |
+
return False, f"β Failed to configure Gemini API: {str(e)}"
|
31 |
+
|
32 |
def cleanup(self):
|
33 |
+
"""Clean up temporary directories and files"""
|
34 |
try:
|
35 |
if hasattr(self, 'download_dir') and os.path.exists(self.download_dir):
|
36 |
shutil.rmtree(self.download_dir)
|
37 |
+
print(f"β
Cleaned up temporary directory: {self.download_dir}")
|
38 |
if hasattr(self, 'temp_downloads') and os.path.exists(self.temp_downloads):
|
39 |
shutil.rmtree(self.temp_downloads)
|
40 |
+
print(f"β
Cleaned up temp downloads directory: {self.temp_downloads}")
|
41 |
except Exception as e:
|
42 |
+
print(f"β οΈ Warning: Could not clean up temporary directory: {e}")
|
43 |
|
44 |
def is_valid_youtube_url(self, url):
|
45 |
youtube_regex = re.compile(
|
|
|
49 |
return youtube_regex.match(url) is not None
|
50 |
|
51 |
def generate_scene_breakdown_gemini(self, video_info):
|
52 |
+
"""Generate AI-powered scene breakdown using Gemini"""
|
53 |
+
if not self.gemini_model:
|
54 |
+
return self.generate_scene_breakdown_fallback(video_info)
|
55 |
+
|
56 |
+
try:
|
57 |
+
duration = video_info.get('duration', 0)
|
58 |
+
title = video_info.get('title', '')
|
59 |
+
description = video_info.get('description', '')[:1000] # Limit description length
|
60 |
+
|
61 |
+
if not duration:
|
62 |
+
return ["**[Duration Unknown]**: Unable to generate timestamped breakdown - video duration not available"]
|
63 |
+
|
64 |
+
# Create prompt for Gemini
|
65 |
+
prompt = f"""
|
66 |
+
Analyze this YouTube video and create a detailed scene-by-scene breakdown with timestamps:
|
67 |
+
|
68 |
+
Title: {title}
|
69 |
+
Duration: {duration} seconds
|
70 |
+
Description: {description}
|
71 |
+
|
72 |
+
Please provide a scene breakdown with the following format:
|
73 |
+
- Divide the video into logical segments based on typical content flow
|
74 |
+
- For videos under 2 minutes: 10-15 second segments
|
75 |
+
- For videos 2-10 minutes: 30-45 second segments
|
76 |
+
- For videos over 10 minutes: 60-90 second segments
|
77 |
+
- Maximum 15 scenes total
|
78 |
+
|
79 |
+
For each scene, provide:
|
80 |
+
**[START_TIME-END_TIME]**: Detailed description of what likely happens in this segment, including visual elements, audio cues, potential dialogue or narration, and scene transitions.
|
81 |
+
|
82 |
+
Consider the video type (tutorial, music video, vlog, etc.) and provide contextually appropriate descriptions.
|
83 |
+
Format timestamps as MM:SS.
|
84 |
+
"""
|
85 |
+
|
86 |
+
response = self.gemini_model.generate_content(prompt)
|
87 |
+
|
88 |
+
# Parse the response into individual scenes
|
89 |
+
if response and response.text:
|
90 |
+
scenes = []
|
91 |
+
lines = response.text.split('\n')
|
92 |
+
for line in lines:
|
93 |
+
line = line.strip()
|
94 |
+
if line and ('**[' in line or line.startswith('*')):
|
95 |
+
scenes.append(line)
|
96 |
+
|
97 |
+
return scenes if scenes else self.generate_scene_breakdown_fallback(video_info)
|
98 |
+
else:
|
99 |
+
return self.generate_scene_breakdown_fallback(video_info)
|
100 |
+
|
101 |
+
except Exception as e:
|
102 |
+
print(f"Gemini API error: {e}")
|
103 |
+
return self.generate_scene_breakdown_fallback(video_info)
|
104 |
|
105 |
+
def generate_scene_breakdown_fallback(self, video_info):
|
106 |
+
"""Fallback scene generation when Gemini is not available"""
|
107 |
+
duration = video_info.get('duration', 0)
|
108 |
+
title = video_info.get('title', '').lower()
|
109 |
+
|
110 |
if not duration:
|
111 |
+
return ["**[Duration Unknown]**: Unable to generate timestamped breakdown"]
|
112 |
+
|
113 |
+
# Simple fallback logic
|
114 |
+
if duration <= 120:
|
115 |
+
segment_length = 15
|
116 |
+
elif duration <= 600:
|
117 |
+
segment_length = 45
|
118 |
+
else:
|
119 |
+
segment_length = 90
|
120 |
+
|
121 |
+
scenes = []
|
122 |
+
num_segments = min(duration // segment_length + 1, 15)
|
123 |
+
|
124 |
+
for i in range(num_segments):
|
125 |
+
start_time = i * segment_length
|
126 |
+
end_time = min(start_time + segment_length - 1, duration)
|
127 |
+
|
128 |
+
start_formatted = f"{start_time//60}:{start_time%60:02d}"
|
129 |
+
end_formatted = f"{end_time//60}:{end_time%60:02d}"
|
130 |
+
|
131 |
+
if i == 0:
|
132 |
+
desc = "Opening sequence with introduction and setup"
|
133 |
+
elif i == num_segments - 1:
|
134 |
+
desc = "Conclusion with final thoughts and call-to-action"
|
135 |
+
else:
|
136 |
+
desc = f"Main content segment {i} with key information and details"
|
137 |
+
|
138 |
+
scenes.append(f"**[{start_formatted}-{end_formatted}]**: {desc}")
|
139 |
+
|
140 |
+
return scenes
|
141 |
|
142 |
+
def detect_video_type(self, title, description):
|
143 |
+
"""Detect video type based on title and description"""
|
144 |
+
text = (title + " " + description).lower()
|
145 |
+
|
146 |
+
if any(word in text for word in ['music', 'song', 'album', 'artist', 'band', 'lyrics']):
|
147 |
+
return "π΅ Music Video"
|
148 |
+
elif any(word in text for word in ['tutorial', 'how to', 'guide', 'learn', 'teaching']):
|
149 |
+
return "π Tutorial/Educational"
|
150 |
+
elif any(word in text for word in ['funny', 'comedy', 'entertainment', 'vlog', 'challenge']):
|
151 |
+
return "π Entertainment/Comedy"
|
152 |
+
elif any(word in text for word in ['news', 'breaking', 'report', 'update']):
|
153 |
+
return "π° News/Information"
|
154 |
+
elif any(word in text for word in ['review', 'unboxing', 'test', 'comparison']):
|
155 |
+
return "β Review/Unboxing"
|
156 |
+
elif any(word in text for word in ['commercial', 'ad', 'brand', 'product']):
|
157 |
+
return "πΊ Commercial/Advertisement"
|
158 |
+
else:
|
159 |
+
return "π¬ General Content"
|
160 |
|
161 |
+
def detect_background_music(self, video_info):
|
162 |
+
"""Detect background music style"""
|
163 |
+
title = video_info.get('title', '').lower()
|
164 |
+
description = video_info.get('description', '').lower()
|
165 |
+
|
166 |
+
if any(word in title for word in ['music', 'song', 'soundtrack']):
|
167 |
+
return "π΅ Original Music/Soundtrack - Primary audio content"
|
168 |
+
elif any(word in title for word in ['commercial', 'ad', 'brand']):
|
169 |
+
return "πΆ Upbeat Commercial Music - Designed to enhance brand appeal"
|
170 |
+
elif any(word in title for word in ['tutorial', 'how to', 'guide']):
|
171 |
+
return "π Minimal/No Background Music - Focus on instruction"
|
172 |
+
elif any(word in title for word in ['vlog', 'daily', 'life']):
|
173 |
+
return "πΌ Ambient Background Music - Complementary to narration"
|
174 |
+
else:
|
175 |
+
return "π΅ Background Music - Complementing video mood and pacing"
|
176 |
|
177 |
+
def detect_influencer_status(self, video_info):
|
178 |
+
"""Detect influencer status"""
|
179 |
+
subscriber_count = video_info.get('channel_followers', 0)
|
180 |
+
view_count = video_info.get('view_count', 0)
|
181 |
+
|
182 |
+
if subscriber_count > 10000000:
|
183 |
+
return "π Mega Influencer (10M+ subscribers)"
|
184 |
+
elif subscriber_count > 1000000:
|
185 |
+
return "β Major Influencer (1M+ subscribers)"
|
186 |
+
elif subscriber_count > 100000:
|
187 |
+
return "π― Mid-tier Influencer (100K+ subscribers)"
|
188 |
+
elif subscriber_count > 10000:
|
189 |
+
return "π Micro Influencer (10K+ subscribers)"
|
190 |
+
elif view_count > 100000:
|
191 |
+
return "π₯ Viral Content Creator"
|
192 |
+
else:
|
193 |
+
return "π€ Regular Content Creator"
|
194 |
|
195 |
+
def format_number(self, num):
|
196 |
+
if num is None or num == 0:
|
197 |
+
return "0"
|
198 |
+
if num >= 1_000_000_000:
|
199 |
+
return f"{num/1_000_000_000:.1f}B"
|
200 |
+
elif num >= 1_000_000:
|
201 |
+
return f"{num/1_000_000:.1f}M"
|
202 |
+
elif num >= 1_000:
|
203 |
+
return f"{num/1_000:.1f}K"
|
204 |
+
return str(num)
|
205 |
|
206 |
def format_video_info(self, video_info):
|
207 |
+
"""Streamlined video information formatting"""
|
208 |
+
if not video_info:
|
209 |
+
return "β No video information available."
|
210 |
+
|
211 |
+
# Basic information
|
212 |
title = video_info.get("title", "Unknown")
|
213 |
uploader = video_info.get("uploader", "Unknown")
|
214 |
duration = video_info.get("duration", 0)
|
|
|
217 |
like_count = video_info.get("like_count", 0)
|
218 |
comment_count = video_info.get("comment_count", 0)
|
219 |
upload_date = video_info.get("upload_date", "Unknown")
|
220 |
+
|
221 |
+
# Format upload date
|
222 |
+
if len(upload_date) == 8:
|
223 |
+
formatted_date = f"{upload_date[:4]}-{upload_date[4:6]}-{upload_date[6:8]}"
|
224 |
+
else:
|
225 |
+
formatted_date = upload_date
|
226 |
|
227 |
+
# Generate enhanced analysis
|
228 |
scene_descriptions = self.generate_scene_breakdown_gemini(video_info)
|
229 |
+
video_type = self.detect_video_type(title, video_info.get('description', ''))
|
230 |
+
background_music = self.detect_background_music(video_info)
|
231 |
+
influencer_status = self.detect_influencer_status(video_info)
|
232 |
+
|
233 |
+
# Calculate engagement metrics
|
234 |
+
engagement_rate = (like_count / view_count) * 100 if view_count > 0 else 0
|
235 |
+
|
236 |
+
# Generate streamlined report
|
237 |
+
report = f"""
|
238 |
+
π¬ YOUTUBE VIDEO ANALYSIS REPORT
|
239 |
+
{'='*50}
|
240 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
241 |
π BASIC INFORMATION
|
242 |
+
{'β'*25}
|
243 |
+
πΉ **Title:** {title}
|
244 |
+
π€ **Uploader:** {uploader}
|
245 |
+
π
**Upload Date:** {formatted_date}
|
246 |
+
β±οΈ **Duration:** {duration_str}
|
247 |
+
π **Video ID:** {video_info.get('id', 'Unknown')}
|
248 |
+
|
249 |
+
π PERFORMANCE METRICS
|
250 |
+
{'β'*25}
|
251 |
+
π **Views:** {self.format_number(view_count)} ({view_count:,})
|
252 |
+
π **Likes:** {self.format_number(like_count)} ({like_count:,})
|
253 |
+
π¬ **Comments:** {self.format_number(comment_count)} ({comment_count:,})
|
254 |
+
π **Engagement Rate:** {engagement_rate:.2f}%
|
255 |
+
|
256 |
+
π― CONTENT ANALYSIS
|
257 |
+
{'β'*25}
|
258 |
+
π **Video Type:** {video_type}
|
259 |
+
π΅ **Background Music:** {background_music}
|
260 |
+
π **Creator Status:** {influencer_status}
|
261 |
+
|
262 |
+
π¬ DETAILED SCENE BREAKDOWN
|
263 |
{'β'*30}
|
264 |
{chr(10).join(scene_descriptions)}
|
265 |
+
|
266 |
+
π DESCRIPTION PREVIEW
|
267 |
+
{'β'*25}
|
268 |
+
{video_info.get('description', 'No description available')[:500]}
|
269 |
+
{'...(truncated)' if len(video_info.get('description', '')) > 500 else ''}
|
270 |
+
|
271 |
+
{'='*50}
|
272 |
+
π **Analysis completed:** {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
|
273 |
+
π€ **AI Enhancement:** {'Gemini AI' if self.gemini_model else 'Standard Analysis'}
|
274 |
"""
|
275 |
+
|
276 |
+
return report.strip()
|
277 |
|
278 |
+
def get_video_info(self, url, progress=gr.Progress(), cookiefile=None):
|
279 |
+
"""Extract video information"""
|
280 |
+
if not url or not url.strip():
|
281 |
+
return None, "β Please enter a YouTube URL"
|
282 |
+
|
283 |
+
if not self.is_valid_youtube_url(url):
|
284 |
+
return None, "β Invalid YouTube URL format"
|
285 |
+
|
286 |
+
try:
|
287 |
+
progress(0.1, desc="Initializing YouTube extractor...")
|
288 |
+
|
289 |
+
ydl_opts = {
|
290 |
+
'noplaylist': True,
|
291 |
+
'extract_flat': False,
|
292 |
+
}
|
293 |
+
|
294 |
+
if cookiefile and os.path.exists(cookiefile):
|
295 |
+
ydl_opts['cookiefile'] = cookiefile
|
296 |
+
|
297 |
+
progress(0.5, desc="Extracting video metadata...")
|
298 |
+
|
299 |
+
with yt_dlp.YoutubeDL(ydl_opts) as ydl:
|
300 |
+
info = ydl.extract_info(url, download=False)
|
301 |
+
|
302 |
+
progress(1.0, desc="β
Analysis complete!")
|
303 |
+
|
304 |
+
return info, "β
Video information extracted successfully"
|
305 |
+
|
306 |
+
except Exception as e:
|
307 |
+
return None, f"β Error: {str(e)}"
|
308 |
+
|
309 |
+
def download_video(self, url, quality="best", audio_only=False, progress=gr.Progress(), cookiefile=None):
|
310 |
+
"""Download video with progress tracking"""
|
311 |
+
if not url or not url.strip():
|
312 |
+
return None, "β Please enter a YouTube URL"
|
313 |
+
|
314 |
+
if not self.is_valid_youtube_url(url):
|
315 |
+
return None, "β Invalid YouTube URL format"
|
316 |
+
|
317 |
+
try:
|
318 |
+
progress(0.1, desc="Preparing download...")
|
319 |
+
|
320 |
+
# Create unique filename
|
321 |
+
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
|
322 |
+
|
323 |
+
# Download to temp directory first (Gradio compatible)
|
324 |
+
ydl_opts = {
|
325 |
+
'outtmpl': os.path.join(self.temp_downloads, f'%(title)s_{timestamp}.%(ext)s'),
|
326 |
+
'noplaylist': True,
|
327 |
+
}
|
328 |
+
|
329 |
+
if audio_only:
|
330 |
+
ydl_opts['format'] = 'bestaudio/best'
|
331 |
+
ydl_opts['postprocessors'] = [{
|
332 |
+
'key': 'FFmpegExtractAudio',
|
333 |
+
'preferredcodec': 'mp3',
|
334 |
+
'preferredquality': '192',
|
335 |
+
}]
|
336 |
+
else:
|
337 |
+
if quality == "best":
|
338 |
+
ydl_opts['format'] = 'best[height<=1080]'
|
339 |
+
elif quality == "720p":
|
340 |
+
ydl_opts['format'] = 'best[height<=720]'
|
341 |
+
elif quality == "480p":
|
342 |
+
ydl_opts['format'] = 'best[height<=480]'
|
343 |
+
else:
|
344 |
+
ydl_opts['format'] = 'best'
|
345 |
+
|
346 |
+
if cookiefile and os.path.exists(cookiefile):
|
347 |
+
ydl_opts['cookiefile'] = cookiefile
|
348 |
+
|
349 |
+
# Progress hook
|
350 |
+
def progress_hook(d):
|
351 |
+
if d['status'] == 'downloading':
|
352 |
+
if 'total_bytes' in d:
|
353 |
+
percent = (d['downloaded_bytes'] / d['total_bytes']) * 100
|
354 |
+
progress(0.1 + (percent / 100) * 0.7, desc=f"Downloading... {percent:.1f}%")
|
355 |
+
else:
|
356 |
+
progress(0.5, desc="Downloading...")
|
357 |
+
elif d['status'] == 'finished':
|
358 |
+
progress(0.8, desc="Processing download...")
|
359 |
+
|
360 |
+
ydl_opts['progress_hooks'] = [progress_hook]
|
361 |
+
|
362 |
+
with yt_dlp.YoutubeDL(ydl_opts) as ydl:
|
363 |
+
info = ydl.extract_info(url, download=True)
|
364 |
+
|
365 |
+
progress(0.9, desc="Copying to Downloads folder...")
|
366 |
+
|
367 |
+
# Find the downloaded file in temp directory
|
368 |
+
downloaded_file_temp = None
|
369 |
+
|
370 |
+
for file in os.listdir(self.temp_downloads):
|
371 |
+
if timestamp in file:
|
372 |
+
downloaded_file_temp = os.path.join(self.temp_downloads, file)
|
373 |
+
break
|
374 |
+
|
375 |
+
if not downloaded_file_temp:
|
376 |
+
return None, "β Downloaded file not found in temp directory"
|
377 |
+
|
378 |
+
# Copy to user's Downloads folder
|
379 |
+
final_filename = os.path.basename(downloaded_file_temp)
|
380 |
+
final_path = os.path.join(self.downloads_folder, final_filename)
|
381 |
+
|
382 |
+
try:
|
383 |
+
shutil.copy2(downloaded_file_temp, final_path)
|
384 |
+
copy_success = True
|
385 |
+
except Exception as e:
|
386 |
+
print(f"Warning: Could not copy to Downloads folder: {e}")
|
387 |
+
copy_success = False
|
388 |
+
final_path = "File downloaded to temp location only"
|
389 |
+
|
390 |
+
progress(1.0, desc="β
Download complete!")
|
391 |
+
|
392 |
+
success_msg = f"""β
Download successful!
|
393 |
+
π Temp file (for download): {os.path.basename(downloaded_file_temp)}
|
394 |
+
π Permanent location: {final_path if copy_success else 'Copy failed'}
|
395 |
+
π― File size: {os.path.getsize(downloaded_file_temp) / (1024*1024):.1f} MB"""
|
396 |
+
|
397 |
+
return downloaded_file_temp, success_msg
|
398 |
+
|
399 |
+
except Exception as e:
|
400 |
+
return None, f"β Download failed: {str(e)}"
|
401 |
+
|
402 |
+
# Initialize global downloader
|
403 |
downloader = YouTubeDownloader()
|
404 |
|
405 |
+
def configure_api_key(api_key):
|
406 |
+
"""Configure Gemini API key"""
|
407 |
+
if not api_key or not api_key.strip():
|
408 |
+
return "β Please enter a valid Google API key", gr.update(visible=False)
|
409 |
+
|
410 |
+
success, message = downloader.configure_gemini(api_key.strip())
|
411 |
+
|
412 |
+
if success:
|
413 |
+
return message, gr.update(visible=True)
|
414 |
+
else:
|
415 |
+
return message, gr.update(visible=False)
|
416 |
+
|
417 |
+
def analyze_with_cookies(url, cookies_file, progress=gr.Progress()):
|
418 |
+
"""Main analysis function"""
|
419 |
+
try:
|
420 |
+
progress(0.05, desc="Starting analysis...")
|
421 |
+
|
422 |
+
cookiefile = None
|
423 |
+
if cookies_file and os.path.exists(cookies_file):
|
424 |
+
cookiefile = cookies_file
|
425 |
+
|
426 |
+
info, msg = downloader.get_video_info(url, progress=progress, cookiefile=cookiefile)
|
427 |
+
|
428 |
+
if info:
|
429 |
+
progress(0.95, desc="Generating comprehensive report...")
|
430 |
+
formatted_info = downloader.format_video_info(info)
|
431 |
+
progress(1.0, desc="β
Complete!")
|
432 |
+
return formatted_info
|
433 |
+
else:
|
434 |
+
return f"β Analysis Failed: {msg}"
|
435 |
+
|
436 |
+
except Exception as e:
|
437 |
+
return f"β System Error: {str(e)}"
|
438 |
|
439 |
+
def download_with_cookies(url, quality, audio_only, cookies_file, progress=gr.Progress()):
|
440 |
+
"""Main download function"""
|
441 |
+
try:
|
442 |
+
progress(0.05, desc="Preparing download...")
|
443 |
+
|
444 |
+
cookiefile = None
|
445 |
+
if cookies_file and os.path.exists(cookies_file):
|
446 |
+
cookiefile = cookies_file
|
447 |
+
|
448 |
+
file_path, msg = downloader.download_video(url, quality, audio_only, progress=progress, cookiefile=cookiefile)
|
449 |
+
|
450 |
+
if file_path:
|
451 |
+
return file_path, msg
|
452 |
+
else:
|
453 |
+
return None, msg
|
454 |
+
|
455 |
+
except Exception as e:
|
456 |
+
return None, f"β System Error: {str(e)}"
|
457 |
+
|
458 |
+
def create_interface():
|
459 |
+
"""Create and configure the Gradio interface"""
|
460 |
+
with gr.Blocks(theme=gr.themes.Soft(), title="π₯ YouTube Video Analyzer & Downloader Pro") as interface:
|
461 |
+
|
462 |
+
gr.HTML("<h1>π₯ YouTube Video Analyzer & Downloader Pro</h1>")
|
463 |
+
|
464 |
+
# API Key Configuration Section
|
465 |
+
with gr.Group():
|
466 |
+
gr.HTML("<h3>π Google Gemini API Configuration</h3>")
|
467 |
+
with gr.Row():
|
468 |
+
api_key_input = gr.Textbox(
|
469 |
+
label="π Google API Key",
|
470 |
+
placeholder="Enter your Google API Key for enhanced AI analysis...",
|
471 |
+
type="password",
|
472 |
+
value=""
|
473 |
+
)
|
474 |
+
configure_btn = gr.Button("π§ Configure API", variant="secondary")
|
475 |
+
|
476 |
+
api_status = gr.Textbox(
|
477 |
+
label="API Status",
|
478 |
+
value="β Gemini API not configured - Using fallback analysis",
|
479 |
+
interactive=False,
|
480 |
+
lines=1
|
481 |
+
)
|
482 |
+
|
483 |
+
# Main Interface (initially hidden until API is configured)
|
484 |
+
main_interface = gr.Group(visible=False)
|
485 |
+
|
486 |
+
with main_interface:
|
487 |
+
with gr.Row():
|
488 |
+
url_input = gr.Textbox(
|
489 |
+
label="π YouTube URL",
|
490 |
+
placeholder="Paste your YouTube video URL here...",
|
491 |
+
value=""
|
492 |
+
)
|
493 |
+
|
494 |
+
cookies_input = gr.File(
|
495 |
+
label="πͺ Upload cookies.txt (Optional)",
|
496 |
+
file_types=[".txt"],
|
497 |
+
type="filepath"
|
498 |
+
)
|
499 |
+
|
500 |
+
with gr.Tabs():
|
501 |
+
with gr.TabItem("π Video Analysis"):
|
502 |
+
analyze_btn = gr.Button("π Analyze Video", variant="primary")
|
503 |
+
|
504 |
+
analysis_output = gr.Textbox(
|
505 |
+
label="π Analysis Report",
|
506 |
+
lines=25,
|
507 |
+
show_copy_button=True
|
508 |
+
)
|
509 |
+
|
510 |
+
analyze_btn.click(
|
511 |
+
fn=analyze_with_cookies,
|
512 |
+
inputs=[url_input, cookies_input],
|
513 |
+
outputs=analysis_output,
|
514 |
+
show_progress=True
|
515 |
+
)
|
516 |
+
|
517 |
+
with gr.TabItem("β¬οΈ Video Download"):
|
518 |
+
with gr.Row():
|
519 |
+
quality_dropdown = gr.Dropdown(
|
520 |
+
choices=["best", "720p", "480p"],
|
521 |
+
value="best",
|
522 |
+
label="πΊ Video Quality"
|
523 |
+
)
|
524 |
+
|
525 |
+
audio_only_checkbox = gr.Checkbox(
|
526 |
+
label="π΅ Audio Only (MP3)",
|
527 |
+
value=False
|
528 |
+
)
|
529 |
+
|
530 |
+
download_btn = gr.Button("β¬οΈ Download Video", variant="primary")
|
531 |
+
|
532 |
+
download_status = gr.Textbox(
|
533 |
+
label="π₯ Download Status",
|
534 |
+
lines=5,
|
535 |
+
show_copy_button=True
|
536 |
+
)
|
537 |
+
|
538 |
+
download_file = gr.File(
|
539 |
+
label="π Downloaded File",
|
540 |
+
visible=False
|
541 |
+
)
|
542 |
+
|
543 |
+
def download_and_update(url, quality, audio_only, cookies_file, progress=gr.Progress()):
|
544 |
+
file_path, status = download_with_cookies(url, quality, audio_only, cookies_file, progress)
|
545 |
+
if file_path and os.path.exists(file_path):
|
546 |
+
return status, gr.update(value=file_path, visible=True)
|
547 |
+
else:
|
548 |
+
return status, gr.update(visible=False)
|
549 |
+
|
550 |
+
download_btn.click(
|
551 |
+
fn=download_and_update,
|
552 |
+
inputs=[url_input, quality_dropdown, audio_only_checkbox, cookies_input],
|
553 |
+
outputs=[download_status, download_file],
|
554 |
+
show_progress=True
|
555 |
+
)
|
556 |
+
|
557 |
+
# Configure API key button action
|
558 |
+
configure_btn.click(
|
559 |
+
fn=configure_api_key,
|
560 |
+
inputs=[api_key_input],
|
561 |
+
outputs=[api_status, main_interface]
|
562 |
+
)
|
563 |
+
|
564 |
+
# Always show interface option (for fallback mode)
|
565 |
+
with gr.Row():
|
566 |
+
show_interface_btn = gr.Button("π Use Without Gemini API (Fallback Mode)", variant="secondary")
|
567 |
+
|
568 |
+
def show_fallback_interface():
|
569 |
+
return "β οΈ Using fallback analysis mode", gr.update(visible=True)
|
570 |
+
|
571 |
+
show_interface_btn.click(
|
572 |
+
fn=show_fallback_interface,
|
573 |
+
outputs=[api_status, main_interface]
|
574 |
+
)
|
575 |
+
|
576 |
+
gr.HTML("""
|
577 |
+
<div style="margin-top: 20px; padding: 15px; background-color: #f0f8ff; border-radius: 10px; border-left: 5px solid #4285f4;">
|
578 |
+
<h3>π How to Get Google API Key:</h3>
|
579 |
+
<ol>
|
580 |
+
<li>Go to <a href="https://console.cloud.google.com/" target="_blank">Google Cloud Console</a></li>
|
581 |
+
<li>Create a new project or select an existing one</li>
|
582 |
+
<li>Enable the "Generative Language API"</li>
|
583 |
+
<li>Go to "Credentials" and create an API key</li>
|
584 |
+
<li>Copy the API key and paste it above</li>
|
585 |
+
</ol>
|
586 |
+
<p><strong>β¨ Benefits of using Gemini API:</strong></p>
|
587 |
+
<ul>
|
588 |
+
<li>π€ AI-powered scene descriptions with contextual understanding</li>
|
589 |
+
<li>π― More accurate content type detection</li>
|
590 |
+
<li>π Enhanced analysis based on video content</li>
|
591 |
+
<li>β° Intelligent timestamp segmentation</li>
|
592 |
+
</ul>
|
593 |
+
</div>
|
594 |
+
""")
|
595 |
+
|
596 |
+
return interface
|
597 |
|
598 |
if __name__ == "__main__":
|
599 |
+
demo = create_interface()
|
600 |
+
import atexit
|
601 |
+
atexit.register(downloader.cleanup)
|
602 |
+
demo.launch(debug=True, show_error=True)
|