Function _create_dataset_for_fragments

python/pyarrow/tests/test_dataset.py:1384–1398

(tempdir, chunk_size=None, filesystem=None)

@pytest.mark.parquet
def _create_dataset_for_fragments(tempdir, chunk_size=None, filesystem=None):
    table = pa.table(
        [range(8), [1] * 8, ['a'] * 4 + ['b'] * 4],
        names=['f1', 'f2', 'part']
    )

    path = str(tempdir / "test_parquet_dataset")

    pq.write_to_dataset(table, path,
                        partition_cols=["part"], chunk_size=chunk_size)
    dataset = ds.dataset(
        path, format="parquet", partitioning="hive", filesystem=filesystem
    )

    return table, dataset


@pytest.mark.parquet