nsheaps/test

Testing

Guidelines for testing during the software development lifecycle.

Test Types

| Type | Scope | When to write | | ----------- | ---------------------------- | -------------------------- | | Unit | Individual functions/modules | During implementation | | Integration | Component interactions | After components are built | | E2E | Full user workflows | After integration | | Smoke | Critical path verification | Before deployment |

Workflow

Step 1: Define Test Strategy

Before writing tests, determine:

• What needs to be tested (from the spec)
• What test types are appropriate
• What the project's existing test patterns look like
• What tools/frameworks are already in use

Step 2: Write Tests

• Follow existing test patterns in the codebase
• Test behavior, not implementation details
• Cover happy paths, error paths, and edge cases
• Keep tests independent and deterministic
• Name tests descriptively (what is being tested and expected outcome)

Step 3: Validate Coverage

• Run the full test suite, not just new tests
• Check that new code has adequate coverage
• Verify tests actually catch the bugs they claim to prevent
• Ensure CI runs the same tests as local

Step 4: Maintain Tests

• Update tests when requirements change
• Delete tests for removed functionality
• Keep test execution fast (prefer unit over E2E where possible)

Coverage Expectations

• New code: aim for meaningful coverage of all public interfaces
• Bug fixes: must include a regression test
• Refactors: existing tests should still pass without modification

Anti-Patterns

| Anti-Pattern | Instead | | ----------------------- | ----------------------------------------- | | Testing implementation | Test behavior and outputs | | Flaky tests | Make tests deterministic | | Testing only happy path | Cover errors and edge cases | | Skipping CI validation | CI is the source of truth | | Giant test files | Organize tests to mirror source structure |