app.py

import os
import requests
import numpy as np
import faiss
from PyPDF2 import PdfReader
from sentence_transformers import SentenceTransformer
from langchain.vectorstores import FAISS
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.chains import RetrievalQA
from langchain.prompts import PromptTemplate
from langchain.llms import GroqLLM
import streamlit as st

# Initialize Groq API LLM
llm = GroqLLM(api_key=os.getenv("GROQ_API_KEY"))

# Function to extract content from a public Google Drive PDF link
def extract_pdf_content(drive_url):
    file_id = drive_url.split("/d/")[1].split("/view")[0]
    download_url = f"https://drive.google.com/uc?export=download&id={file_id}"
    response = requests.get(download_url)
    if response.status_code != 200:
        return None

    with open("document.pdf", "wb") as f:
        f.write(response.content)

    reader = PdfReader("document.pdf")
    text = ""
    for page in reader.pages:
        text += page.extract_text()
    return text

# Function to create a FAISS vector store from the document content
def create_vector_store(text):
    sentences = text.split(". ")
    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    vector_store = FAISS.from_texts(sentences, embedding=embeddings)
    return vector_store, sentences

# Streamlit app
st.title("RAG-based Application with Focused Context")

# Predefined Google Drive link
drive_url = "https://drive.google.com/file/d/1XvqA1OIssRs2gbmOtKFKj-02yQ5X2yg0/view?usp=sharing"

# Extract document content
st.write("Extracting content from the document...")
text = extract_pdf_content(drive_url)
if text:
    st.write("Document extracted successfully!")

    st.write("Creating vector store...")
    vector_store, sentences = create_vector_store(text)

    st.write("Vector store created successfully!")

    query = st.text_input("Enter your query:")
    if query:
        st.write("Retrieving relevant context from the document...")
        retriever = vector_store.as_retriever()
        retriever.search_kwargs["k"] = 3  # Retrieve top 3 matches

        # Define a prompt template to guide LLM response generation
        prompt_template = PromptTemplate(
            template="""
            Use the following context to answer the question:
            
            {context}
            
            Question: {question}
            Answer:""",
            input_variables=["context", "question"]
        )

        # Create a RetrievalQA chain
        qa_chain = RetrievalQA(
            retriever=retriever,
            llm=llm,
            prompt=prompt_template
        )

        # Run the query through the QA chain
        result = qa_chain.run(query)
        st.write("Answer:", result)
else:
    st.error("Failed to extract content from the document.")