import gradio as gr
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.chat_models import ChatOpenAI
from langchain.chains import ConversationalRetrievalChain
from langchain.vectorstores import FAISS
from langchain.document_loaders.csv_loader import CSVLoader

def conversational_chat(query, lang_model_key, file_upload):
    global chain, session_history
    if file_upload is not None:
        # Loading the CSV file
        loader = CSVLoader(file_path=file_upload, encoding="utf-8")
        data = loader.load()

        # Initializing embeddings and vectors
        embeddings = OpenAIEmbeddings(openai_api_key=lang_model_key)
        vectors = FAISS.from_documents(data, embeddings)

        # Creating the ConversationalRetrievalChain
        chain = ConversationalRetrievalChain.from_llm(llm=ChatOpenAI(temperature=0.0, model_name='gpt-4', openai_api_key=lang_model_key),
                                                      retriever=vectors.as_retriever())
        session_history = []

    result = chain({"question": query, "chat_history": session_history})
    session_history.append((query, result["answer"]))
    return result["answer"]

iface = gr.Interface(fn=conversational_chat,
                     inputs=[gr.Textbox(label="Query", lines=7),
                             gr.Textbox(label="Your OpenAI API key:", type="password"),
                             gr.File(label="Upload your CSV file:", type="binary")],
                     outputs="text",
                     title="Conversational CSV Chat: Please upload your file and set your API Key in order to use the functionalities",

                     )
iface.launch()