Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -391,38 +391,43 @@ def process_youtube_video(url="", keywords=""):
|
|
| 391 |
if not video_id:
|
| 392 |
return thumbnail, "Invalid YouTube URL", sentiment_label, recommendations
|
| 393 |
|
| 394 |
-
video_id = extract_video_id(url)
|
| 395 |
-
if not video_id:
|
| 396 |
-
return thumbnail, "Invalid YouTube URL", sentiment_label, recommendations
|
| 397 |
-
|
| 398 |
-
# Set thumbnail
|
| 399 |
thumbnail = f"https://img.youtube.com/vi/{video_id}/maxresdefault.jpg"
|
| 400 |
|
| 401 |
try:
|
| 402 |
-
#
|
| 403 |
-
|
| 404 |
-
|
| 405 |
-
|
| 406 |
-
|
| 407 |
-
|
| 408 |
try:
|
| 409 |
-
|
| 410 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 411 |
except:
|
| 412 |
-
|
|
|
|
|
|
|
|
|
|
| 413 |
|
| 414 |
-
|
| 415 |
-
transcript = transcript_list.find_generated_transcript(['en'])
|
| 416 |
-
|
| 417 |
-
# Get transcript text
|
| 418 |
-
text = " ".join([t['text'] for t in transcript.fetch()])
|
| 419 |
-
|
| 420 |
-
# Clean text
|
| 421 |
cleaned_text = re.sub(r'[^\w\s.]', '', text)
|
| 422 |
cleaned_text = ' '.join(cleaned_text.split())
|
| 423 |
|
| 424 |
# Sentiment Analysis
|
| 425 |
-
blob = TextBlob(cleaned_text[:2000])
|
| 426 |
polarity = blob.sentiment.polarity
|
| 427 |
subjectivity = blob.sentiment.subjectivity
|
| 428 |
|
|
@@ -436,10 +441,9 @@ def process_youtube_video(url="", keywords=""):
|
|
| 436 |
model = genai.GenerativeModel("gemini-pro")
|
| 437 |
summary = model.generate_content(f"Summarize this content: {cleaned_text[:4000]}").text
|
| 438 |
|
| 439 |
-
except (TranscriptsDisabled, NoTranscriptFound):
|
| 440 |
-
return thumbnail, "⚠️ No English subtitles available", "N/A", recommendations
|
| 441 |
except Exception as e:
|
| 442 |
-
|
|
|
|
| 443 |
|
| 444 |
# Get recommendations
|
| 445 |
if keywords.strip():
|
|
@@ -448,6 +452,7 @@ def process_youtube_video(url="", keywords=""):
|
|
| 448 |
return thumbnail, summary, sentiment_label, recommendations
|
| 449 |
|
| 450 |
except Exception as e:
|
|
|
|
| 451 |
return None, f"Error: {str(e)}", "N/A", ""
|
| 452 |
|
| 453 |
|
|
|
|
| 391 |
if not video_id:
|
| 392 |
return thumbnail, "Invalid YouTube URL", sentiment_label, recommendations
|
| 393 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 394 |
thumbnail = f"https://img.youtube.com/vi/{video_id}/maxresdefault.jpg"
|
| 395 |
|
| 396 |
try:
|
| 397 |
+
# Method 1: Direct transcript fetch
|
| 398 |
+
try:
|
| 399 |
+
transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=['en'])
|
| 400 |
+
text = " ".join([t['text'] for t in transcript])
|
| 401 |
+
except:
|
| 402 |
+
# Method 2: Try list_transcripts
|
| 403 |
try:
|
| 404 |
+
transcript_list = YouTubeTranscriptApi.list_transcripts(video_id)
|
| 405 |
+
|
| 406 |
+
# Try multiple language variants
|
| 407 |
+
for lang_code in ['en', 'en-US', 'en-GB', 'a.en']:
|
| 408 |
+
try:
|
| 409 |
+
transcript = transcript_list.find_transcript([lang_code])
|
| 410 |
+
text = " ".join([t['text'] for t in transcript.fetch()])
|
| 411 |
+
break
|
| 412 |
+
except:
|
| 413 |
+
continue
|
| 414 |
+
|
| 415 |
+
# If no English transcript found, try auto-generated
|
| 416 |
+
if 'text' not in locals():
|
| 417 |
+
transcript = transcript_list.find_generated_transcript(['en'])
|
| 418 |
+
text = " ".join([t['text'] for t in transcript.fetch()])
|
| 419 |
except:
|
| 420 |
+
# Method 3: Try translation
|
| 421 |
+
available_transcripts = transcript_list.find_manually_created_transcript()
|
| 422 |
+
translated = available_transcripts.translate('en')
|
| 423 |
+
text = " ".join([t['text'] for t in translated.fetch()])
|
| 424 |
|
| 425 |
+
# Clean and process text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 426 |
cleaned_text = re.sub(r'[^\w\s.]', '', text)
|
| 427 |
cleaned_text = ' '.join(cleaned_text.split())
|
| 428 |
|
| 429 |
# Sentiment Analysis
|
| 430 |
+
blob = TextBlob(cleaned_text[:2000])
|
| 431 |
polarity = blob.sentiment.polarity
|
| 432 |
subjectivity = blob.sentiment.subjectivity
|
| 433 |
|
|
|
|
| 441 |
model = genai.GenerativeModel("gemini-pro")
|
| 442 |
summary = model.generate_content(f"Summarize this content: {cleaned_text[:4000]}").text
|
| 443 |
|
|
|
|
|
|
|
| 444 |
except Exception as e:
|
| 445 |
+
print(f"Debug - Transcript Error: {str(e)}") # Debug logging
|
| 446 |
+
return thumbnail, f"⚠️ Unable to process video: {str(e)}", "N/A", recommendations
|
| 447 |
|
| 448 |
# Get recommendations
|
| 449 |
if keywords.strip():
|
|
|
|
| 452 |
return thumbnail, summary, sentiment_label, recommendations
|
| 453 |
|
| 454 |
except Exception as e:
|
| 455 |
+
print(f"Debug - Main Error: {str(e)}") # Debug logging
|
| 456 |
return None, f"Error: {str(e)}", "N/A", ""
|
| 457 |
|
| 458 |
|