app.py

import os
import streamlit as st
import pickle
import time
import requests
from bs4 import BeautifulSoup
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain.chains import RetrievalQAWithSourcesChain
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.vectorstores import Chroma
from langchain_groq import ChatGroq
from dotenv import load_dotenv
from langchain.schema import Document
from langchain.vectorstores import FAISS
from langchain_openai import OpenAI
from langchain.llms import OpenAI


load_dotenv()  # Load environment variables from .env file

st.title("RockyBot: News Research Tool 📈")
st.sidebar.title("News Article URLs")

# Collect URLs from user input
urls = [st.sidebar.text_input(f"URL {i+1}") for i in range(3)]
process_url_clicked = st.sidebar.button("Process URLs")
file_path = "faiss_store_openai.pkl"

main_placeholder = st.empty()
llm = OpenAI(model_name="gpt-4o", temperature=0.9, max_tokens=500)
#llm = ChatGroq(model_name="llama-3.3-70b-versatile", temperature=0.9, max_tokens=500)

def fetch_web_content(url):
    """Fetches text content from a given URL using BeautifulSoup."""
    try:
        response = requests.get(url, timeout=10)
        response.raise_for_status()
        soup = BeautifulSoup(response.text, "html.parser")
        return soup.get_text()
    except Exception as e:
        return f"Error fetching {url}: {str(e)}"

if process_url_clicked:
    main_placeholder.text("Data Loading...Started...✅✅✅")
    
    # Fetch content from URLs
    data = [fetch_web_content(url) for url in urls if url.strip()]
    
    main_placeholder.text("Data Loading...Completed...✅✅✅")
    
    # Split data into chunks
    text_splitter = RecursiveCharacterTextSplitter(
        separators=['\n\n', '\n', '.', ','],
        chunk_size=1000
    )
    main_placeholder.text("Text Splitting...Started...✅✅✅")
    docs = [Document(page_content=text) for text in data]
    docs = text_splitter.split_documents(docs)
    #docs = text_splitter.split_documents(data)
    
    # Create embeddings and save to Chroma vector store
    embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    #vectorstore_huggingface = Chroma.from_documents(docs, embedding_model)
    vectorstore_huggingface = FAISS.from_documents(docs, embedding_model)
    main_placeholder.text("Embedding Vector Started Building...✅✅✅")
    time.sleep(2)
    
    # Save the vector store to a pickle file
    with open(file_path, "wb") as f:
        pickle.dump(vectorstore_huggingface, f)

# User query input
query = st.text_input("Question: ")
if query:
    if os.path.exists(file_path):
        with open(file_path, "rb") as f:
            vectorstore = pickle.load(f)
            chain = RetrievalQAWithSourcesChain.from_llm(llm=llm, retriever=vectorstore.as_retriever())
            result = chain({"question": query}, return_only_outputs=True)
            
            # Display answer
            st.header("Answer")
            st.write(result["answer"])
            
            # Display sources, if available
            sources = result.get("sources", "")
            if sources:
                st.subheader("Sources:")
                sources_list = sources.split("\n")
                for source in sources_list:
                    st.write(source)