Instruction file imported from rcarmo/go-busybox (.github/instructions/testing.instructions.md). Copyright stays with the author.

Testing instructions

Applies when: writing or modifying tests in Go projects.

Mandatory: Use Parameterized Tests

All tests with multiple cases MUST use table-driven (parameterized) approach:

func TestFoo(t *testing.T) {
    tests := []struct {
        name string
        // inputs and expected outputs
    }{
        {"case1", ...},
        {"case2", ...},
    }
    for _, tt := range tests {
        t.Run(tt.name, func(t *testing.T) {
            // test logic
        })
    }
}

Use testutil Package

For busybox-wasm applet tests, use the pkg/testutil helpers:

import "github.com/rcarmo/go-busybox/pkg/testutil"

func TestApplet(t *testing.T) {
    tests := []testutil.AppletTestCase{
        {Name: "basic", Args: []string{"arg"}, WantCode: 0, WantOut: "expected\n"},
    }
    testutil.RunAppletTests(t, applet.Run, tests)
}

Coverage Requirements

• Run make coverage before committing
• Minimum 80% coverage for new code
• Critical paths (sandbox, security) require 100%

Duplicate Code

• Run make dupl to check for duplicates
• Extract common patterns to shared helpers
• Use t.Helper() for test helper functions

Test Organization

• Tests live alongside code: foo.go → foo_test.go
• Shared utilities go in pkg/testutil/
• Use t.TempDir() for filesystem tests (auto-cleanup)