Function test_fuzzy_sorting

scripts/test_stringzilla.py:1169–1204

(list_length: int, part_length: int, variability: int, seed_value: int)

@pytest.mark.parametrize("variability", [2, 3])
@pytest.mark.parametrize("seed_value", SEED_VALUES)
def test_fuzzy_sorting(list_length: int, part_length: int, variability: int, seed_value: int):
    seed_random_generators(seed_value)
    native_list = [get_random_string(variability=variability, length=part_length) for _ in range(list_length)]
    native_joined = ".".join(native_list)
    big_joined = Str(native_joined)
    big_list = big_joined.split(".")

    # Before testing sorting, validate pairwise comparator consistency
    def py_cmp(a: str, b: str) -> int:
        return -1 if a < b else (1 if a > b else 0)

    def sz_cmp(a: str, b: str) -> int:
        sa, sb = Str(a), Str(b)
        if sa < sb:
            return -1
        if sa > sb:
            return 1
        return 0

    # Check every consecutive pair a[i], a[i+1]
    for i in range(len(native_list) - 1):
        a, b = native_list[i], native_list[i + 1]
        assert py_cmp(a, b) == sz_cmp(a, b), f"Comparator mismatch at {i}: '{a}' vs '{b}'"

    native_ordered = sorted(native_list)
    native_order = big_list.argsort()
    for i in range(list_length):
        assert native_ordered[i] == native_list[native_order[i]], "Order is wrong"
        assert native_ordered[i] == str(big_list[int(native_order[i])]), "Split is wrong?!"

    native_list.sort()
    big_list = big_list.sorted()

    assert len(native_list) == len(big_list)
    for native_str, big_str in zip(native_list, big_list):
        assert native_str == str(big_str), "Order is wrong"


@pytest.mark.skipif(not pyarrow_available, reason="PyArrow is not installed")