from dotenv import load_dotenv
from langchain_openai import ChatOpenAI
from langchain_core.tools import tool
from datetime import datetime, timezone
from googleapiclient.discovery import build
from app.rag.split_docs import load_split_docs
from app.rag.llm import load_llm_openai
from app.rag.embeddings import load_embeddins
from app.rag.retriever import create_retriever
from app.rag.vectorstore import create_vectorstore
from app.rag.rag_chain import create_rag_chain
import pytz
import telebot
import os


class LangChainTools:
    def load_llm_openai(self):
        load_dotenv()
        # model = "gpt-3.5-turbo-0125"
        # model = "gpt-4o"
        model = "gpt-4o-mini"

        llm = ChatOpenAI(
            model=model,
            temperature=0.1,
            max_tokens=2000,
        )
        return llm


@tool
def redact_email(topic: str) -> str:
    """Use this tool to draft the content of an email based on a topic."""

    # Load LLM model
    langChainTools = LangChainTools()

    llm = langChainTools.load_llm_openai()
    # Create prompt for the LLM
    prompt = (
        "Please redact a email based on the topic:\n\n"
        "Topic: {}\n\n"
        "Email Content: [Your email content here]"
    ).format(topic)

    response = llm.invoke(prompt)
    return response


@tool
def send_message(message: str):
    """Use this function when you need to communicate with Cristian."""
    # Configuración del bot
    load_dotenv()
    API_TOKEN_BOT = os.getenv("API_TOKEN_BOT")
    bot = telebot.TeleBot(API_TOKEN_BOT)

    # Escapar caracteres especiales en Markdown
    from telebot.util import escape_markdown

    safe_message = escape_markdown(message)

    # Enviar mensaje usando MarkdownV2
    bot.send_message(chat_id="5076346205", text=safe_message,
                     parse_mode="Markdown")


@tool
def get_company_info(prompt: str) -> str:
    """
    Use this function when you need more information
    about the services offered by OneCluster.
    """
    file_path: str = "onecluster_info.pdf"

    try:
        docs_split: list = load_split_docs(file_path)
        embeddings_model = load_embeddins()
        llm = load_llm_openai()

        # Usar el nombre corregido de la función
        create_vectorstore(docs_split, embeddings_model, file_path)

        retriever = create_retriever(
            embeddings_model, persist_directory="embeddings/onecluster_info"
        )
        qa = create_rag_chain(llm, retriever)

        response = qa.invoke({"input": prompt, "chat_history": []})

        return response["answer"]
    except Exception as e:
        print(f"Error en get_company_info: {e}")
        return f"Lo siento, hubo un error al procesar la información: {str(e)}"


@tool
def get_current_date_and_time():
    """
    Use this function when you need to know the current date and time.

    Returns:
        str: Current date and time in Bogotá, Colombia.
    """
    bogota_tz = pytz.timezone("America/Bogota")
    current_date_and_time = datetime.now(bogota_tz)
    return current_date_and_time.strftime("%Y-%m-%d %H:%M:%S")