import streamlit as st
from langchain_google_genai import GoogleGenerativeAIEmbeddings, ChatGoogleGenerativeAI
from langchain.vectorstores import FAISS
from langchain.chains.question_answering import load_qa_chain
from langchain.prompts import PromptTemplate
import google.generativeai as genai
from dotenv import load_dotenv
import os

# Load environment variables
load_dotenv()
os.getenv("GOOGLE_API_KEY")
genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))

# Define the function to load the vector store
def load_vector_store():
    embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001")
    # Load the vector store with dangerous deserialization allowed
    vector_store = FAISS.load_local("faiss_index", embeddings, allow_dangerous_deserialization=True)
    return vector_store

# Define the function to get the conversational chain
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 in
    provided context just say, "answer is not available in the context", don't provide the wrong answer\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

# Define the function to handle user input
def handle_user_query(user_question):
    vector_store = load_vector_store()
    docs = vector_store.similarity_search(user_question)

    chain = get_conversational_chain()

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

    return response.get("output_text", "No response generated.")

# Define the main function for Streamlit app
def main():
    st.set_page_config("Chat with PDF")
    st.header("ASK about general theory of relativity")

    # Load the vector store initially
    if 'vector_store' not in st.session_state:
        st.session_state.vector_store = load_vector_store()
    
    # Text input for user query
    user_question = st.text_input("Ask a Question")

    if user_question:
        response = handle_user_query(user_question)
        st.write("Reply:", response)

if __name__ == "__main__":
    main()