Update app.py

app.py

@@ -1,149 +1,158 @@
-import streamlit as st
-from youtube_transcript_api import YouTubeTranscriptApi, TranscriptsDisabled, NoTranscriptFound
-from nltk.tokenize import word_tokenize
-from nltk.corpus import stopwords
-from nltk.stem import WordNetLemmatizer
-from sklearn.feature_extraction.text import CountVectorizer
-from sklearn.decomposition import LatentDirichletAllocation
-from transformers import pipeline
-from textblob import TextBlob
-import re
-import nltk
-
-# Ensure that necessary NLTK data is downloaded
-nltk.download('punkt')
-nltk.download('stopwords')
-nltk.download('wordnet')
-
-# Function to summarize text
-def summarize_text(text, max_length=80000):  # Increased max_length to 80,000
-    summarization_pipeline = pipeline("summarization")
-    summary = summarization_pipeline(text, max_length=max_length, min_length=100, do_sample=False)
-    return summary[0]['summary_text']
-
-# Function to extract keywords
-def extract_keywords(text):
-    stop_words = set(stopwords.words('english'))
-    lemmatizer = WordNetLemmatizer()
-
-    words = word_tokenize(text)
-    words = [lemmatizer.lemmatize(word.lower()) for word in words if word.isalnum()]
-    keywords = [word for word in words if word not in stop_words and len(word) > 1]
-
-    counter = CountVectorizer().fit_transform([' '.join(keywords)])
-    vocabulary = CountVectorizer().fit([' '.join(keywords)]).vocabulary_
-    top_keywords = sorted(vocabulary, key=vocabulary.get, reverse=True)[:5]
-
-    return top_keywords
-
-# Function to perform topic modeling
-def topic_modeling(text):
-    vectorizer = CountVectorizer(max_df=2, min_df=0.95, stop_words='english')
-    tf = vectorizer.fit_transform([text])
-    lda_model = LatentDirichletAllocation(n_components=5, max_iter=5, learning_method='online', random_state=42)
-    lda_model.fit(tf)
-    feature_names = vectorizer.get_feature_names_out()
-    topics = []
-    for topic_idx, topic in enumerate(lda_model.components_):
-        topics.append([feature_names[i] for i in topic.argsort()[:-6:-1]])
-    return topics
-
-# Function to extract YouTube video ID from URL
-def extract_video_id(url):
-    video_id = None
-    patterns = [
-        r'v=([^&]+)',  # Pattern for URLs with 'v=' parameter
-        r'youtu.be/([^?]+)',  # Pattern for shortened URLs
-        r'youtube.com/embed/([^?]+)'  # Pattern for embed URLs
-    ]
-    for pattern in patterns:
-        match = re.search(pattern, url)
-        if match:
-            video_id = match.group(1)
-            break
-    return video_id
-
-# Main Streamlit app
-def main():
-    st.title("YouTube Video Summarizer")
-
-    # Sidebar
-    st.sidebar.title("App Features and Description")
-    st.sidebar.subheader("What does this app do?")
-    st.sidebar.write("""
-    - Extracts the transcript of a YouTube video.
-    - Summarizes the video content.
-    - Extracts keywords from the transcript.
-    - Performs topic modeling.
-    - Conducts sentiment analysis to show how positive or negative the content is.
-    """)
-    
-    st.sidebar.subheader("Sentiment Analysis Explained")
-    st.sidebar.write("""
-    - **Polarity**: Measures how positive or negative the content is. Ranges from -1 to 1. 
-        - Negative polarity (< 0) indicates negative sentiment.
-        - Positive polarity (> 0) indicates positive sentiment.
-        - Neutral polarity (0) indicates neutral sentiment.
-    - **Subjectivity**: Measures how subjective or objective the content is. Ranges from 0 to 1.
-        - A subjectivity score closer to 1 indicates personal opinions or beliefs.
-        - A score closer to 0 indicates factual information.
-    """)
-
-    # User input for YouTube video URL
-    video_url = st.text_input("Enter YouTube Video URL:", "")
-
-    # User customization options
-    max_summary_length = st.slider("Max Summary Length:", 1000, 80000, 50000)  # Increased max length to 80,000
-
-    if st.button("Summarize"):
-        try:
-            # Extract video ID from URL
-            video_id = extract_video_id(video_url)
-            if not video_id:
-                st.error("Invalid YouTube URL. Please enter a valid URL.")
-                return
-
-            # Get transcript of the video
-            transcript = YouTubeTranscriptApi.get_transcript(video_id)
-            if not transcript:
-                st.error("Transcript not available for this video. Please try a different video.")
-                return
-
-            video_text = ' '.join([line['text'] for line in transcript])
-
-            # Summarize the transcript
-            summary = summarize_text(video_text, max_length=max_summary_length)
-
-            # Extract keywords from the transcript
-            keywords = extract_keywords(video_text)
-
-            # Perform topic modeling
-            topics = topic_modeling(video_text)
-
-            # Perform sentiment analysis
-            sentiment = TextBlob(video_text).sentiment
-
-            # Display summarized text, keywords, topics, and sentiment
-            st.subheader("Video Summary:")
-            st.write(summary)
-
-            st.subheader("Keywords:")
-            st.write(keywords)
-
-            st.subheader("Topics:")
-            for idx, topic in enumerate(topics):
-                st.write(f"Topic {idx+1}: {', '.join(topic)}")
-
-            st.subheader("Sentiment Analysis:")
-            st.write(f"Polarity: {sentiment.polarity}")
-            st.write(f"Subjectivity: {sentiment.subjectivity}")
-
-        except TranscriptsDisabled:
-            st.error("Transcripts are disabled for this video. Please try a different video.")
-        except NoTranscriptFound:
-            st.error("No transcript found for this video. Please try a different video.")
-        except Exception as e:
-            st.error(f"Error: {str(e)}")
-
-if __name__ == "__main__":
-    main()
+import streamlit as st
+from youtube_transcript_api import YouTubeTranscriptApi, TranscriptsDisabled, NoTranscriptFound
+from nltk.tokenize import word_tokenize
+from nltk.corpus import stopwords
+from nltk.stem import WordNetLemmatizer
+from sklearn.feature_extraction.text import CountVectorizer
+from sklearn.decomposition import LatentDirichletAllocation
+from transformers import pipeline
+from textblob import TextBlob
+import re
+import nltk
+
+# Ensure that necessary NLTK data is downloaded
+nltk.download('punkt')
+nltk.download('stopwords')
+nltk.download('wordnet')
+
+# Function to summarize text
+def summarize_text(text, max_length=80000):  # Increased max_length to 80,000
+    summarization_pipeline = pipeline("summarization", model="facebook/bart-large-cnn")
+    summary = summarization_pipeline(text, max_length=max_length, min_length=100, do_sample=False)
+    return summary[0]['summary_text']
+
+# Function to extract keywords
+def extract_keywords(text):
+    stop_words = set(stopwords.words('english'))
+    lemmatizer = WordNetLemmatizer()
+
+    words = word_tokenize(text)
+    words = [lemmatizer.lemmatize(word.lower()) for word in words if word.isalnum()]
+    keywords = [word for word in words if word not in stop_words and len(word) > 1]
+
+    counter = CountVectorizer().fit_transform([' '.join(keywords)])
+    vocabulary = CountVectorizer().fit([' '.join(keywords)]).vocabulary_
+    top_keywords = sorted(vocabulary, key=vocabulary.get, reverse=True)[:5]
+
+    return top_keywords
+
+# Function to perform topic modeling
+def topic_modeling(text):
+    vectorizer = CountVectorizer(max_df=2, min_df=0.95, stop_words='english')
+    tf = vectorizer.fit_transform([text])
+    lda_model = LatentDirichletAllocation(n_components=5, max_iter=5, learning_method='online', random_state=42)
+    lda_model.fit(tf)
+    feature_names = vectorizer.get_feature_names_out()
+    topics = []
+    for topic_idx, topic in enumerate(lda_model.components_):
+        topics.append([feature_names[i] for i in topic.argsort()[:-6:-1]])
+    return topics
+
+# Function to extract YouTube video ID from URL
+def extract_video_id(url):
+    video_id = None
+    patterns = [
+        r'v=([^&]+)',  # Pattern for URLs with 'v=' parameter
+        r'youtu.be/([^?]+)',  # Pattern for shortened URLs
+        r'youtube.com/embed/([^?]+)'  # Pattern for embed URLs
+    ]
+    for pattern in patterns:
+        match = re.search(pattern, url)
+        if match:
+            video_id = match.group(1)
+            break
+    return video_id
+
+# Function to fetch transcript with retries
+def get_transcript(video_id):
+    try:
+        # Attempt to fetch the transcript with language preference
+        transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=['en'])
+        return transcript
+    except TranscriptsDisabled:
+        st.error("Transcripts are disabled for this video. Try a different one.")
+    except NoTranscriptFound:
+        st.error("No transcript found for this video. Ensure it has captions enabled.")
+    except Exception as e:
+        st.error(f"An unexpected error occurred while fetching the transcript: {str(e)}")
+    return None
+
+# Main Streamlit app
+def main():
+    st.title("YouTube Video Summarizer")
+
+    # Sidebar
+    st.sidebar.title("App Features and Description")
+    st.sidebar.subheader("What does this app do?")
+    st.sidebar.write("""
+    - Extracts the transcript of a YouTube video.
+    - Summarizes the video content.
+    - Extracts keywords from the transcript.
+    - Performs topic modeling.
+    - Conducts sentiment analysis to show how positive or negative the content is.
+    """)
+    
+    st.sidebar.subheader("Sentiment Analysis Explained")
+    st.sidebar.write("""
+    - **Polarity**: Measures how positive or negative the content is. Ranges from -1 to 1. 
+        - Negative polarity (< 0) indicates negative sentiment.
+        - Positive polarity (> 0) indicates positive sentiment.
+        - Neutral polarity (0) indicates neutral sentiment.
+    - **Subjectivity**: Measures how subjective or objective the content is. Ranges from 0 to 1.
+        - A subjectivity score closer to 1 indicates personal opinions or beliefs.
+        - A score closer to 0 indicates factual information.
+    """)
+
+    # User input for YouTube video URL
+    video_url = st.text_input("Enter YouTube Video URL:", "")
+
+    # User customization options
+    max_summary_length = st.slider("Max Summary Length:", 1000, 80000, 50000)  # Increased max length to 80,000
+
+    if st.button("Summarize"):
+        try:
+            # Extract video ID from URL
+            video_id = extract_video_id(video_url)
+            if not video_id:
+                st.error("Invalid YouTube URL. Please enter a valid URL.")
+                return
+
+            # Get transcript of the video
+            transcript = get_transcript(video_id)
+            if not transcript:
+                return
+
+            video_text = ' '.join([line['text'] for line in transcript])
+
+            # Summarize the transcript
+            summary = summarize_text(video_text, max_length=max_summary_length)
+
+            # Extract keywords from the transcript
+            keywords = extract_keywords(video_text)
+
+            # Perform topic modeling
+            topics = topic_modeling(video_text)
+
+            # Perform sentiment analysis
+            sentiment = TextBlob(video_text).sentiment
+
+            # Display summarized text, keywords, topics, and sentiment
+            st.subheader("Video Summary:")
+            st.write(summary)
+
+            st.subheader("Keywords:")
+            st.write(keywords)
+
+            st.subheader("Topics:")
+            for idx, topic in enumerate(topics):
+                st.write(f"Topic {idx+1}: {', '.join(topic)}")
+
+            st.subheader("Sentiment Analysis:")
+            st.write(f"Polarity: {sentiment.polarity}")
+            st.write(f"Subjectivity: {sentiment.subjectivity}")
+
+        except Exception as e:
+            st.error(f"Error: {str(e)}")
+
+if __name__ == "__main__":
+    main()