app.py

import streamlit as st
from PyPDF2 import PdfReader
from langchain.text_splitter import RecursiveCharacterTextSplitter
import os
from langchain_google_genai import GoogleGenerativeAIEmbeddings
import google.generativeai as genai
from langchain.vectorstores import FAISS
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain.chains.question_answering import load_qa_chain
from langchain.prompts import PromptTemplate
from dotenv import load_dotenv
import secrets


# helper functions

def get_pdf_text(pdf_docs): # function to obtain raw text from pdf
    text=""
    for pdf in pdf_docs:
        pdf_reader= PdfReader(pdf)
        for page in pdf_reader.pages:
            text+= page.extract_text()
    return text

def get_text_chunks(text): # function to slipt text into chucks for efficient and faster processing
    text_split = RecursiveCharacterTextSplitter(chunk_size=10000, chunk_overlap=1000)
    chunks = text_split.split_text(text)
    return chunks

def get_vector_store(text_chunks): # function to get vector store and embedding
    embeddings = GoogleGenerativeAIEmbeddings(model = "models/embedding-001")
    vector_store = FAISS.from_texts(text_chunks, embedding=embeddings)
    vector_store.save_local("faiss_index")

def get_conversational_chain():

    prompt_template = """
    Answer the question as detailed as possible from the provided context, 
    make sure to provide all the details. 
    If the answer is not present within the text, 
    say that the given question cannot be answered, 
    DO NOT make up things on your own.
    Do not return the question in the response, please.\n\n
    Context:\n {context}?\n
    Question: \n{question}\n

    Answer:
    """

    model = ChatGoogleGenerativeAI(model="gemini-pro",
                             temperature=0.3)

    prompt = PromptTemplate(template = prompt_template, input_variables = ["context", "question"])
    chain = load_qa_chain(model, chain_type="stuff", prompt=prompt)

    return chain



def user_input(user_question):
    if not api_key: 
        st.error("Please authenticate first by entering your API key.")
        return
    
    embeddings = GoogleGenerativeAIEmbeddings(model = "models/embedding-001")
    
    new_db = FAISS.load_local("faiss_index", embeddings,allow_dangerous_deserialization=True)
    docs = new_db.similarity_search(user_question)

    chain = get_conversational_chain()

    
    response = chain(
        {"input_documents":docs, "question": user_question}
        , return_only_outputs=True)

    print(response)
    st.write("Gemini: ", response["output_text"])

# main code
st.title("Research Paper Chatbot")
st.write("Chat with any research paper pdf, powered by Gemini 🔹")

with st.sidebar:
    st.title("Gemini API")
    
    api_key = st.text_input("Enter your Gemini API key", type="password")
    if api_key:
        genai.configure(api_key=api_key)
    else:
        if "api_key" in st.secrets:
            genai.configure(api_key=st.secrets["api_key"])
        else:
            st.error("Missing API key.")

    pdf_docs = st.file_uploader("Upload your research paper", accept_multiple_files=True)
    if (pdf_docs):
        st.write("Click Go to start")

    if st.button("Go!"):
        with st.spinner("Embedding document... This may take a while..."):
            raw_text = get_pdf_text(pdf_docs)
            text_chunks = get_text_chunks(raw_text)
            get_vector_store(text_chunks)
            st.success("Ready! Let's chat!")

user_question = st.text_input("Ask a question from the paper")

if user_question:
    user_input(user_question)