property-based-testing

Use the Hypothesis library with the @given decorator and strategies like st.integers() or st.lists() to generate inputs. Define assertions about properties that must hold for all inputs, such as len(sorted(lst)) == len(lst), and Hypothesis generates and shrinks test cases automatically.

Example-based testing checks specific hand-picked inputs like add(2, 3) == 5, while property-based testing asserts invariants like add(a, b) == add(b, a) for all generated inputs. Property tests discover edge cases automatically; example tests document known behavior and regressions.

Shrinking is the process where the framework reduces a failing input to the minimal case that still fails. For example, a failing 150-element list might shrink to [0, -1], making the root cause of the bug immediately visible.

Hypothesis is the standard choice for Python with excellent shrinking and stateful testing support. fast-check serves JavaScript and TypeScript projects with good performance, async property support, and model-based testing commands.

Avoid property-based testing for non-deterministic behavior, tests requiring specific known edge cases, or when generators would be too slow. Keep example-based tests for bug reproductions, documented behavior, and fast deterministic checks alongside property tests.

Reproduce the failure manually using the exact minimal counterexample the framework reports, then trace execution with debug logging. Watch for boundary values like 0, empty collections, or integer limits, which often reveal missing null handling or overflow issues.