PY-TEST0014 · finite_range_hypothesis_candidate_count

Count broad integer ranges that should use a Hypothesis strategy.

This is a deterministic rule for python. Read its implementation.

Definition

Inspect tests collected by Pytest’s default Python conventions. Report pytest.mark.parametrize when its values are a static range containing at least minimum_cases integers. The default is ten. A long contiguous domain is generation, not a short example table, and Hypothesis can explore it while shrinking a failure to a minimal counterexample.

Evidence

Each finding identifies the decorated test and records the exact finite range size. The rule does not claim that every short example table should use property-based testing. The value is the number of broad finite ranges a strategy would state better.

Exceptions

Dynamic ranges, noninteger case sources, short curated boundary tables, and already generated strategies abstain. Keep parametrization when every value has distinct domain meaning or needs an explicit case ID or mark. A Hypothesis test still needs a stated invariant.

Examples

Bad

@pytest.mark.parametrize("value", range(100))
def test_round_trip(value):
    assert decode(encode(value)) == value

Good

@given(st.integers(min_value=0, max_value=99))
def test_round_trip(value):
    assert decode(encode(value)) == value

References

• Cites “Hypothesis documentation”, introduction and when to use property-based testing. Open reference
• Cites “Hypothesis documentation”, quickstart. Open reference