helper_function.py

import io
import streamlit as st
from PyPDF2 import PdfReader
from dotenv import load_dotenv
from langchain_groq import ChatGroq
from langchain_community.vectorstores import FAISS
from langchain.memory import ConversationBufferMemory
from langchain.chains import ConversationalRetrievalChain
from langchain.text_splitter import CharacterTextSplitter
from langchain_community.embeddings import HuggingFaceInstructEmbeddings

from PyPDF2 import PdfReader
import io

from PyPDF2 import PdfReader
import io

def get_pdf_text(pdf_docs):
    text = ""
    for pdf in pdf_docs:
    
        pdf_reader = PdfReader(io.BytesIO(pdf))
        for page in pdf_reader.pages:
            text += page.extract_text() or "" 
    return text


def get_text_chunks(text):
    text_splitter = CharacterTextSplitter(
        separator="\n",
        chunk_size=1000,
        chunk_overlap=200,
        length_function=len
    )
    chunks = text_splitter.split_text(text)
    return chunks

def get_vectorstore(text_chunks):
    embeddings = HuggingFaceInstructEmbeddings(model_name="all-MiniLM-L12-v2")
    vectorstore = FAISS.from_texts(texts=text_chunks, embedding=embeddings)
    return vectorstore

def get_conversation_chain(vectorstore):
    llm = ChatGroq(model="gemma2-9b-it")
    memory = ConversationBufferMemory(
        memory_key='chat_history', return_messages=True
    )
    conversation_chain = ConversationalRetrievalChain.from_llm(
        llm=llm,
        retriever=vectorstore.as_retriever(),
        memory=memory
    )
    return conversation_chain



def handle_userinput(user_question):
    if 'conversation' not in st.session_state:
        st.error("Conversation not initialized. Please upload and process PDF documents first.")
        return

    conversation_chain = st.session_state.conversation

    # Process user input using the appropriate method
    response = conversation_chain.run({'question': user_question})
    
    final_answer = response.get('answer', 'Sorry, I couldn\'t find an answer.')
    st.markdown(f"**Response:** {final_answer}")
    st.markdown("---")