import random
import string

import pandas as pd
from pandas.api.types import is_bool_dtype, is_numeric_dtype

from nicegui import ui


def generate_random_dataframe(*, range_columns=(5, 8), range_rows=(5, 10)):
    """Returns a pandas dataframe with a random number of rows and columns, each
    containing random data of various data types.
    """

    # Generate random number of columns
    num_columns = random.randint(*range_columns)

    # Generate random column names
    column_names = [''.join(random.choices(string.ascii_letters, k=random.randint(3, 10))) for i in range(num_columns)]

    # Generate random data types for each column
    column_types = [random.choice([int, float, str, bool]) for i in range(num_columns)]

    # Generate random data for each row
    rows = []
    num_rows = random.randint(*range_rows)
    for i in range(num_rows):
        row = []
        for j in range(num_columns):
            if column_types[j] == int:
                row.append(random.randint(-100, 100))
            elif column_types[j] == float:
                row.append(random.uniform(-100, 100))
            elif column_types[j] == str:
                row.append(''.join(random.choices(string.ascii_letters, k=random.randint(3, 10))))
            elif column_types[j] == bool:
                row.append(random.choice([True, False]))
        rows.append(row)

    # Add a column with integers and the same name as another column
    column_name = random.choice(column_names)
    column_names.append(column_name + ' ID')
    column_types.append(int)
    for row in rows:
        row.append(random.randint(1, 100))

    # Create the dataframe
    return pd.DataFrame(rows, columns=column_names)


def update(*, df: pd.DataFrame, r: int, c: int, value):
    df.iat[r, c] = value
    ui.notify(f"Set ({r}, {c}) to {value}")


@ui.refreshable
def home():
    df = generate_random_dataframe()
    with ui.header():
        ui.button("Randomize", on_click=home.refresh)
    with ui.row():
        for c, col in enumerate(df.columns):
            with ui.column():
                ui.label(col.capitalize())
                for r, row in enumerate(df.loc[:, col]):
                    with ui.row().classes("h-8 items-center"):
                        if is_numeric_dtype(df[col].dtype):
                            cls = ui.number
                        elif is_bool_dtype(df[col].dtype):
                            cls = ui.checkbox
                        else:
                            cls = ui.input
                        cls(value=row, on_change=lambda event, r=r, c=c: update(df=df, r=r, c=c, value=event.value))


home()
ui.run(native=True)