go-testing-code-review

Table-driven tests in Go use a slice of test cases with input and expected output fields, executed in a loop with t.Run for each case. This pattern improves readability, reduces code duplication, and makes it easy to add new test scenarios by simply adding rows to the table.

Use t.Run subtests with t.Parallel to run test cases concurrently within table-driven patterns. Each subtest receives its own copy of loop variables, avoiding race conditions and speeding up test execution across multiple cores.

Register cleanup functions with t.Cleanup instead of defer statements. This ensures teardown runs in reverse order of registration, handles panics safely, and isolates cleanup logic from test setup, improving reliability and maintainability.

Clear, descriptive error messages in assertions help identify exactly what failed and why when tests break. Combined with table-driven test case names, they reduce debugging time and make test failures immediately actionable.

Yes, interface-based mocks work well in table-driven tests by storing mock instances in test case structs. This approach isolates mock behavior per case, avoids shared mutable state, and keeps test cases independent and reproducible.