import logging
import re

import pandas as pd

from openbb_terminal.decorators import log_start_end
from openbb_terminal.helper_funcs import get_user_agent, request

logger = logging.getLogger(__name__)


def format_number(text: str) -> float:
    numbers = re.findall(r"[-+]?(?:\d*\.\d+|\d+)", text)
    if "t" in text.lower():
        multiplier = 1_000_000_000_000
    elif "b" in text.lower():
        multiplier = 1_000_000_000
    elif "m" in text.lower():
        multiplier = 1_000_000
    else:
        multiplier = 1

    return round(float(numbers[0]) * multiplier, 2)


@log_start_end(log=logger)
def get_debt() -> pd.DataFrame:
    """Retrieves national debt information for various countries. [Source: wikipedia.org]

    Returns
    -------
    pd.DataFrame
        Country, Debt [$], Debt [% of GDP], Debt per capita [$], Debt per capita [% of GDP]
    """
    url = "https://en.wikipedia.org/wiki/List_of_countries_by_external_debt"
    response = request(url, headers={"User-Agent": get_user_agent()})
    df = pd.read_html(response.content)[0]
    df = df.rename(
        columns={
            "Country/Region": "Country",
            "External debt US dollars": "USD Debt",
            "Per capitaUS dollars": "Per Capita",
            "Per capita US dollars": "USD Per Capita",
            "Date": "As Of",
        }
    )
    df["USD Debt"] = df["USD Debt"].apply(lambda x: format_number(x)).astype("int64")
    df["USD Per Capita"] = df["USD Per Capita"].astype("int64")
    df["Rank"] = df["USD Debt"].rank(ascending=False).astype("int64")
    date = df["As Of"].astype(str)
    dates = []
    for i in date.index:
        dates.append(date[i].split("[")[0])
    df["As Of"] = dates
    df["As Of"] = df["As Of"].str.replace("31 September", "30 September").str.strip()
    indexes = ["Rank", "As Of", "Country", "USD Per Capita", "USD Debt", "% of GDP"]
    df = df[indexes]

    return df