tests/conftest.py

"""Shared fixtures for the data_cleaning_env test suite."""

from __future__ import annotations

import pandas as pd
import pytest


@pytest.fixture
def clean_df() -> pd.DataFrame:
    """Minimal clean DataFrame for unit tests."""
    return pd.DataFrame(
        {
            "a": [1, 2, 3, 4, 5],
            "b": ["x", "y", "z", "x", "y"],
            "c": [10.0, 20.0, 30.0, 40.0, 50.0],
        }
    )


@pytest.fixture
def dirty_df(clean_df: pd.DataFrame) -> pd.DataFrame:
    """Dirty version of clean_df with a missing value and a type error."""
    import numpy as np

    df = clean_df.copy()
    df.loc[0, "a"] = np.nan
    df["c"] = df["c"].astype(object)
    df.loc[2, "c"] = "err_30.0"
    return df