import os
import time
import streamlit as st
from htmlTemplates import css, bot_template, user_template
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import Chroma
from langchain.memory import ConversationBufferMemory
from langchain.chains import RetrievalQA
from pdfminer.high_level import extract_text
from langchain.text_splitter import RecursiveCharacterTextSplitter
from transformers import AutoTokenizer, AutoModelForCausalLM


# Updated Prompt Template

tokenizer = AutoTokenizer.from_pretrained("TinyPixel/Llama-2-7B-bf16-sharded")
model = AutoModelForCausalLM.from_pretrained("TinyPixel/Llama-2-7B-bf16-sharded")

persist_directory = 'db'
embeddings_model_name = 'sentence-transformers/all-MiniLM-L6-v2'

def get_pdf_text(pdf_path):
    return extract_text(pdf_path)

def get_pdf_text_chunks(pdf_text):
    text_splitter = RecursiveCharacterTextSplitter()
    return text_splitter.split_text(text=pdf_text, max_chunk_length=1000, min_chunk_length=100, overlap_length=100)

def create_vector_store(target_source_chunks):
    embeddings = HuggingFaceEmbeddings(model_name=embeddings_model_name)
    db = Chroma(persist_directory=persist_directory, embedding_function=embeddings)
    db.add(target_source_chunks)
    return db

def get_vector_store(target_source_chunks):
    embeddings = HuggingFaceEmbeddings(model_name=embeddings_model_name)
    db = Chroma(persist_directory=persist_directory, embedding_function=embeddings)
    retriver = db.as_retriever(search_kwargs={"k": target_source_chunks})
    return retriver

def get_conversation_chain(retriever):
    memory = ConversationBufferMemory(memory_key='chat_history', return_messages=True,)
    chain = RetrievalQA.from_llm(
        llm=model,
        memory=memory,
        retriever=retriever,
    )
    return chain


def handle_userinput(user_question):
    if st.session_state.conversation is None:
        st.warning("Please load the Vectorstore first!")
        return
    else:
        with st.spinner('Thinking...', ):
            start_time = time.time()
            response = st.session_state.conversation({'query': user_question})
            end_time = time.time()

            st.session_state.chat_history = response['chat_history']

            for i, message in enumerate(st.session_state.chat_history):
                if i % 2 == 0:
                    st.write(user_template.replace("{{MSG}}", message.content), unsafe_allow_html=True)
                else:
                    st.write(bot_template.replace("{{MSG}}", message.content), unsafe_allow_html=True)

            st.write('Elapsed time: {:.2f} seconds'.format(end_time - start_time))
            st.balloons()




def main():

    st.set_page_config(page_title='Chat with PDF', page_icon=':rocket:', layout='wide', )
    with st.sidebar.title(':gear: Parameters'):
        model_n_ctx = st.sidebar.slider('Model N_CTX', min_value=128, max_value=2048, value=1024, step=2)
        model_n_batch = st.sidebar.slider('Model N_BATCH', min_value=1, max_value=model_n_ctx, value=512, step=2)
        target_source_chunks = st.sidebar.slider('Target Source Chunks', min_value=1, max_value=10, value=4, step=1)
    st.write(css, unsafe_allow_html=True)

    if "conversation" not in st.session_state:
        st.session_state.conversation = None
    if "chat_history" not in st.session_state:
        st.session_state.chat_history = None

    st.header('Chat with PDF :robot_face:')
    st.subheader('Upload your PDF file and start chatting with it!')
    user_question = st.text_input('Enter your message here:')

    if st.button('Start Chain'):
        with st.spinner('Working in progress ...'):
            pdf_file = st.file_uploader("Upload PDF", type=['pdf'])
            if pdf_file is not None:
                pdf_text = get_pdf_text(pdf_file)
                pdf_text_chunks = get_pdf_text_chunks(pdf_text)
                st.session_state.vector_store = create_vector_store(pdf_text_chunks)
                st.session_state.conversation = get_conversation_chain(
                    retriever=st.session_state.vector_store,
                )
                st.success('Vectorstore created successfully! You can start chatting now!')
            else:
                st.warning('Please upload a PDF file first!')


    if user_question:
        handle_userinput(user_question)


if __name__ == '__main__':
    main()