import os

import qdrant_client
import streamlit as st
from langchain.chains import RetrievalQA
from langchain.llms import HuggingFaceHub
from langchain_community.vectorstores import FAISS
from langchain_openai import OpenAIEmbeddings

client = qdrant_client.QdrantClient(
    os.getenv("qdrant_host"), api_key=os.getenv("qdrant_key")
)


def main():
    # st.set_page_config(page_title="Chat with multiple PDFs",
    #                   page_icon=":books:")
    # st.write(css, unsafe_allow_html=True)

    st.set_page_config(page_title="Ask Qdrant", page_icon=":books:")
    st.header("Ask your remote database 💬")

    embeddings = OpenAIEmbeddings()

    db = FAISS.load_local("faiss_index", embeddings)

    llm = HuggingFaceHub(
        repo_id="bigscience/bloom",
        model_kwargs={"temperature": 0.2, "max_length": 512, "max_new_tokens": 100},
    )

    qa = RetrievalQA.from_chain_type(
        llm=llm, chain_type="stuff", retriever=db.as_retriever()
    )

    # show user input
    user_question = st.text_input("Ask a question Mastercard's available APIs:")
    if user_question:
        answer = qa.invoke(user_question)
        st.write(f"Question: {answer['query']}")
        st.write(f"Answer: {answer['result']}")

    col1, col2, col3 = st.columns([1, 6, 1])

    with col1:
        st.write("")

    with col2:
        st.write("")

    with col3:
        st.image("mc_symbol_opt_73_3x.png")


if __name__ == "__main__":
    main()