app.py

import os
import validators
import streamlit as st
from dotenv import load_dotenv
from langchain.prompts import PromptTemplate
from langchain_groq import ChatGroq
from langchain.chains.summarize import load_summarize_chain
from langchain.schema import Document
from youtube_transcript_api import YouTubeTranscriptApi
from urllib.parse import urlparse, parse_qs
import requests
from bs4 import BeautifulSoup
import re

# Load environment variables
load_dotenv()

# Streamlit App
st.set_page_config(page_title="AI Content Summarizer", page_icon="📚")

# Create two columns for the title
col1, col2 = st.columns([0.85, 0.15])
with col1:
    st.title("AI Content Summarizer")
    st.caption("Powered by LangChain & Gemma 🤖")
    
with col2:
    st.image("https://python.langchain.com/img/favicon.ico", width=50)

st.markdown("""
### About This App
This application leverages the power of LangChain and Gemma AI to automatically generate concise summaries from YouTube videos and web articles. Whether you're researching a topic, catching up on content, or trying to quickly grasp key information, this tool can help save time by distilling content into clear, readable summaries.
""")

# Get API Key & URL input
groq_api_key = os.getenv("GROQ_API_KEY")
if not groq_api_key:
    st.error("GROQ API Key not found. Please check your environment variables.")

generic_url = st.text_input("Enter YouTube or Website URL", placeholder="https://example.com or https://youtube.com/watch?v=...")

# LangChain Model with Groq API
llm = ChatGroq(model="gemma-7b-it", groq_api_key=groq_api_key)

# Prompt Template
prompt_template = """
Provide a clear and concise summary in 300 words of the following content:

{text}

Focus on the main points and key insights. Write in a professional tone.
"""
prompt = PromptTemplate(template=prompt_template, input_variables=["text"])

def get_youtube_id(url):
    """Extract video ID from YouTube URL"""
    if 'youtube.com' in url:
        query = parse_qs(urlparse(url).query)
        return query.get('v', [None])[0]
    elif 'youtu.be' in url:
        return urlparse(url).path[1:]
    return None

def get_youtube_content(url):
    """Get content from YouTube video using youtube-transcript-api"""
    try:
        video_id = get_youtube_id(url)
        if not video_id:
            raise ValueError("Could not extract YouTube video ID")

        # Get transcript
        transcript = YouTubeTranscriptApi.get_transcript(video_id)
        transcript_text = ' '.join([entry['text'] for entry in transcript])

        # Get video info using a simple request
        response = requests.get(f"https://www.youtube.com/oembed?url=https://www.youtube.com/watch?v={video_id}&format=json")
        if response.status_code == 200:
            video_info = response.json()
            title = video_info.get('title', 'Unknown Title')
            author = video_info.get('author_name', 'Unknown Author')
        else:
            title = "Unknown Title"
            author = "Unknown Author"

        content = f"""
Video Title: {title}
Channel: {author}
Transcript:
{transcript_text}
"""
        return [Document(page_content=content)]

    except Exception as e:
        st.error(f"Error getting YouTube content: {str(e)}")
        return None

def get_website_content(url):
    """Get content from website using requests and BeautifulSoup"""
    try:
        headers = {
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'
        }
        response = requests.get(url, headers=headers, verify=False)
        response.raise_for_status()
        
        soup = BeautifulSoup(response.text, 'html.parser')
        
        # Remove unwanted elements
        for element in soup(['script', 'style', 'header', 'footer', 'nav']):
            element.decompose()
            
        title = soup.title.string if soup.title else "No title found"
        
        # Get main content with improved selection
        main_content = ""
        selectors = [
            'article', 'main', 
            '[role="main"]', 
            '.post-content', 
            '.article-content', 
            '.content',
            '#content'
        ]
        
        for selector in selectors:
            if soup.select_one(selector):
                main_content = soup.select_one(selector).get_text()
                break
        
        if not main_content:
            paragraphs = soup.find_all('p')
            main_content = '\n'.join(p.get_text() for p in paragraphs)
        
        # Clean up text
        main_content = re.sub(r'\s+', ' ', main_content).strip()
        main_content = re.sub(r'<[^>]+>', '', main_content)
        
        content = f"""
Title: {title}
URL: {url}

Content:
{main_content}
"""
        return [Document(page_content=content)]
        
    except Exception as e:
        st.error(f"Error processing website content: {str(e)}")
        return None

if st.button("Summarize"):
    if not groq_api_key or not generic_url.strip():
        st.error("Please provide a valid API key and URL.")
    elif not validators.url(generic_url):
        st.error("Please enter a valid URL.")
    else:
        try:
            with st.spinner("Processing content..."):
                if "youtube.com" in generic_url or "youtu.be" in generic_url:
                    docs = get_youtube_content(generic_url)
                else:
                    docs = get_website_content(generic_url)
                
                if docs is None:
                    st.stop()

                chain = load_summarize_chain(llm, chain_type="stuff", prompt=prompt)
                output_summary = chain.run(docs)

                st.success("Summary Generated!")
                
                tab1, tab2 = st.tabs(["Summary", "Original Content"])
                
                with tab1:
                    st.markdown(output_summary)
                    
                with tab2:
                    if docs:
                        st.text_area("Raw Content", docs[0].page_content, height=300)

        except Exception as e:
            st.error(f"An error occurred: {str(e)}")