import pandas as pd
import pytest
from src.cleaner.dataframe import build_df


@pytest.fixture
def sample_objs():
    return [
        {'id': 1, 'name': 'Alice', 'age': 25},
        {'id': 2, 'name': 'Bob', 'age': 30},
        {'id': 3, 'name': 'Charlie', 'age': 22}
    ]


def test_build_df(sample_objs):
    result_df = build_df(sample_objs)

    assert isinstance(result_df, pd.DataFrame)
    assert set(result_df.columns) == {'name', 'age'}


def test_build_df_missing_id_column():
    objs_missing_id = [{'name': 'Alice', 'age': 25}, {'name': 'Bob', 'age': 30}]
    with pytest.raises(KeyError, match="'id'"):
        build_df(objs_missing_id)