Test Smells

Test smells are indicators of problems in test code — patterns that make tests fragile, unreadable, or unreliable.

Mystery Guest — the test uses external resources, and thus is not self-contained.
Resource Optimism — the test makes assumptions about the existence/absence of state from external resources.
Test Run War — test code fails when multiple programmers are running them (shared state collisions).
General Fixture — a fixture is over-general: it sets up more than any one test needs, making the tests harder to understand.
Eager Test — a single test checks too many methods at once. Solution: separate into tests that each check one method.
Lazy Test — several tests check the same method using the same fixture (duplication with little additional signal).
Assertion Roulette — many assertions in one test, with no descriptive messages, making it hard to know which one failed.
Indirect Testing — a test exercises its target class only through another class, so failures are hard to attribute.
For Testers Only — the production class exposes methods that exist only to support testing.
Sensitive Equality — the test relies on toString or similar fragile equality that breaks when formatting changes.
Test Code Duplication — the same test logic is copy-pasted across multiple tests instead of being factored out.

🛠️ Steven Gong