Testing

Go Testing

Go has a built-in test framework — no external library needed. Files end in _test.go, functions start with Test, and go test runs everything.

Always run go test -race ./... in CI — the race detector catches data races that otherwise ship to production.

Test File Conventions

Table-Driven Tests

The canonical Go testing pattern:

func TestParseSize(t *testing.T) {
    tests := []struct {
        name    string
        input   string
        want    int64
        wantErr bool
    }{
        {name: "bytes", input: "100B", want: 100},
        {name: "kilobytes", input: "2KB", want: 2048},
        {name: "megabytes", input: "1MB", want: 1048576},
        {name: "empty", input: "", wantErr: true},
        {name: "invalid", input: "abc", wantErr: true},
    }

    for _, tt := range tests {
        t.Run(tt.name, func(t *testing.T) {
            got, err := ParseSize(tt.input)
            if tt.wantErr {
                if err == nil {
                    t.Fatal("expected error, got nil")
                }
                return
            }
            if err != nil {
                t.Fatalf("unexpected error: %v", err)
            }
            if got != tt.want {
                t.Errorf("ParseSize(%q) = %d, want %d", tt.input, got, tt.want)
            }
        })
    }
}

Testing Decision Table

t.Parallel() Loop-Variable Capture

Before Go 1.22, a parallel subtest in a range loop captures the loop variable by reference — all subtests see the final iteration's value. This is the same root cause covered in Closures; pin it per iteration:

for _, tt := range tests {
    tt := tt // pin (unnecessary on Go 1.22+, where range vars are per-iteration)
    t.Run(tt.name, func(t *testing.T) {
        t.Parallel()
        // ...
    })
}

Anti-patterns

Read On Demand

Benchmark

Scenario: .benchmarks/scenarios/golang-testing-001-table-driven.md · Run: 2026-08-31 · Log: .benchmarks/runs/2026-08-31/golang-testing-001-table-driven.json

NEUTRAL (run 2026-08-31). All models 100% with and without — table-driven tests are default behavior at this task size. Gate per .agents/skills/skill-optimizer/rules/release-gates.md.