Property-Based Testing
Table of Contents
Overview
Property-based testing verifies that code satisfies general properties or invariants for a wide range of automatically generated inputs, rather than testing specific examples. This approach finds edge cases and bugs that example-based tests often miss.
When to Use
- Testing algorithms with mathematical properties
- Verifying invariants that should always hold
- Finding edge cases automatically
- Testing parsers and serializers (round-trip properties)
- Validating data transformations
- Testing sorting, searching, and data structure operations
- Discovering unexpected input combinations
Quick Start
Minimal working example:
# test_string_operations.py
import pytest
from hypothesis import given, strategies as st, assume, example
def reverse_string(s: str) -> str:
"""Reverse a string."""
return s[::-1]
class TestStringOperations:
@given(st.text())
def test_reverse_twice_returns_original(self, s):
"""Property: Reversing twice returns the original string."""
assert reverse_string(reverse_string(s)) == s
@given(st.text())
def test_reverse_length_unchanged(self, s):
"""Property: Reverse doesn't change length."""
assert len(reverse_string(s)) == len(s)
@given(st.text(min_size=1))
def test_reverse_first_becomes_last(self, s):
"""Property: First char becomes last after reverse."""
reversed_s = reverse_string(s)
assert s[0] == reversed_s[-1]
assert s[-1] == reversed_s[0]
// ... (see reference guides for full implementation)
Reference Guides
Detailed implementations in the references/ directory:
| Guide | Contents |
|---|---|
| Hypothesis for Python | Hypothesis for Python |
| fast-check for JavaScript/TypeScript | fast-check for JavaScript/TypeScript |
| junit-quickcheck for Java | junit-quickcheck for Java |
Best Practices
✅ DO
- Focus on general properties, not specific cases
- Test mathematical properties (commutativity, associativity)
- Verify round-trip encoding/decoding
- Use shrinking to find minimal failing cases
- Combine with example-based tests for known edge cases
- Test invariants that should always hold
- Generate realistic input distributions
❌ DON'T
- Test properties that are tautologies
- Over-constrain input generation
- Ignore shrunk test failures
- Replace all example tests with properties
- Test implementation details
- Generate invalid inputs without constraints
- Forget to handle edge cases in generators