Testing Infrastructure

Testing Infrastructure

Test Runners

Vitest

Primary test runner for:

• Unit tests (.test.ts, .test.tsx)
• Storybook story tests (via addon-vitest)

Running Tests

# Unit tests only
pnpm test

# Specific package
pnpm test --filter @lsst-sqre/squared

# Storybook tests
pnpm test-storybook
pnpm test-storybook:watch
pnpm test-storybook --filter @lsst-sqre/squared

# Comprehensive CI pipeline
pnpm run localci  # Or use test-suite-runner agent

Vitest Configuration

Squared Package

Two test projects in vitest.config.ts:

Unit tests:

• Files: src/**/*.test.{ts,tsx}
• Environment: jsdom
• Setup: src/test-setup.ts
• CSS Modules: non-scoped strategy

Storybook tests:

• Browser: Playwright (chromium)
• Environment: browser + jsdom
• Setup: .storybook/vitest.setup.ts
• Plugin: @storybook/addon-vitest

Writing Unit Tests

Basic Pattern

import { describe, it, expect, vi } from 'vitest';
import { render, screen } from '@testing-library/react';
import userEvent from '@testing-library/user-event';
import MyComponent from './MyComponent';

describe('MyComponent', () => {
  it('renders correctly', () => {
    render(<MyComponent title="Test" />);
    expect(screen.getByText('Test')).toBeInTheDocument();
  });

  it('handles user interaction', async () => {
    const handleClick = vi.fn();
    render(<MyComponent />);

    await userEvent.click(screen.getByRole('button'));
    expect(handleClick).toHaveBeenCalled();
  });
});

React Testing Library

Query priority:

1. getByRole - Most accessible
2. getByLabelText - For forms
3. getByPlaceholderText - For inputs
4. getByText - For content
5. getByTestId - Last resort

// Good - semantic queries
screen.getByRole('button', { name: 'Submit' });
screen.getByLabelText('Email address');

// Avoid - implementation details
container.querySelector('.button');

User Interactions

import userEvent from '@testing-library/user-event';

it('handles form submission', async () => {
  const user = userEvent.setup();
  render(<MyForm />);

  await user.type(screen.getByLabelText('Email'), 'test@example.com');
  await user.click(screen.getByRole('button', { name: 'Submit' }));

  expect(handleSubmit).toHaveBeenCalledWith({
    email: 'test@example.com',
  });
});

Storybook Tests

In Stories

import { expect, within, userEvent } from '@storybook/test';

export const WithInteraction: Story = {
  tags: ['test'],
  play: async ({ canvasElement }) => {
    const canvas = within(canvasElement);

    // Assert initial state
    expect(canvas.getByText('Click me')).toBeInTheDocument();

    // Interact
    await userEvent.click(canvas.getByRole('button'));

    // Assert result
    expect(canvas.getByText('Clicked!')).toBeInTheDocument();
  },
};

Running Story Tests

# Run all story tests
pnpm test-storybook --filter @lsst-sqre/squared

# Watch mode
pnpm test-storybook:watch --filter @lsst-sqre/squared

# Specific story
pnpm test-storybook --filter @lsst-sqre/squared -- --grep "MyComponent"

Mocking

Mock Functions

import { vi } from 'vitest';

const mockFn = vi.fn();
const mockFnWithReturn = vi.fn(() => 'result');

expect(mockFn).toHaveBeenCalledWith('arg');
expect(mockFn).toHaveBeenCalledTimes(2);

Mock Modules

vi.mock('../api/client', () => ({
  fetchData: vi.fn(() => Promise.resolve({ data: 'mock' })),
}));

Mock SWR

import useSWR from 'swr';

vi.mock('swr');

it('renders with data', () => {
  vi.mocked(useSWR).mockReturnValue({
    data: { name: 'Test' },
    error: undefined,
    isLoading: false,
    isValidating: false,
    mutate: vi.fn(),
  });

  render(<MyComponent />);
  expect(screen.getByText('Test')).toBeInTheDocument();
});

Mock Data

Mock Data Patterns

// src/lib/mocks/userData.ts
export const mockUserData = {
  username: 'testuser',
  email: 'test@example.com',
  groups: ['admin'],
};

export const mockUserDataLoading = {
  data: undefined,
  error: undefined,
  isLoading: true,
};

export const mockUserDataError = {
  data: undefined,
  error: new Error('Failed to fetch'),
  isLoading: false,
};

MSW (Mock Service Worker)

For API mocking in Storybook:

// .storybook/preview.tsx
import { initialize, mswLoader } from 'msw-storybook-addon';
import { handlers } from '../src/mocks/handlers';

initialize({ onUnhandledRequest: 'bypass' });

export const loaders = [mswLoader];
export const parameters = {
  msw: {
    handlers,
  },
};

// src/mocks/handlers.ts
import { http, HttpResponse } from 'msw';

export const handlers = [
  http.get('/api/user-info', () => {
    return HttpResponse.json({
      username: 'testuser',
      email: 'test@example.com',
    });
  }),
];

Testing Configuration Components

import { AppConfigProvider } from '../contexts/AppConfigContext';

const mockConfig = {
  siteName: 'Test Site',
  baseUrl: 'http://localhost:3000',
  // ... other required config
};

function renderWithConfig(ui: React.ReactElement) {
  return render(
    <AppConfigProvider config={mockConfig}>
      {ui}
    </AppConfigProvider>
  );
}

it('uses config', () => {
  renderWithConfig(<MyComponent />);
  expect(screen.getByText('Test Site')).toBeInTheDocument();
});

Async Testing

Waiting for Elements

// Wait for element to appear
await screen.findByText('Loaded data');

// Wait for element to disappear
await waitForElementToBeRemoved(() => screen.queryByText('Loading...'));

// Custom wait
await waitFor(() => {
  expect(screen.getByText('Done')).toBeInTheDocument();
});

Testing Async Functions

it('fetches data', async () => {
  const { result } = renderHook(() => useMyData());

  // Wait for loading to complete
  await waitFor(() => {
    expect(result.current.isLoading).toBe(false);
  });

  expect(result.current.data).toEqual(expectedData);
});

Coverage

# Run tests with coverage
pnpm test -- --coverage

# Coverage thresholds in vitest.config.ts
test: {
  coverage: {
    reporter: ['text', 'json', 'html'],
    thresholds: {
      lines: 80,
      functions: 80,
      branches: 80,
      statements: 80,
    },
  },
}

Debugging Tests

# Run single test file
pnpm test --filter @lsst-sqre/squared -- MyComponent.test.tsx

# Run in watch mode
pnpm test --filter @lsst-sqre/squared -- --watch

# Debug in browser (for Storybook tests)
pnpm test-storybook:watch --filter @lsst-sqre/squared

Debug Queries

import { screen } from '@testing-library/react';

// Print DOM tree
screen.debug();

// Print specific element
screen.debug(screen.getByRole('button'));

// Log available roles
screen.logTestingPlaygroundURL();

CI Pipeline

test-suite-runner Agent

Use the test-suite-runner agent (via Task tool) for comprehensive testing:

// Agent runs: pnpm run localci
// Which includes:
// - pnpm format:check
// - pnpm lint
// - pnpm type-check
// - pnpm test
// - pnpm build

The agent:

• Runs full CI pipeline
• Analyzes failures across all stages
• Provides detailed failure reports
• Suggests fixes

Best Practices

1. Test behavior, not implementation
2. Use semantic queries (getByRole, etc.)
3. Mock external dependencies (APIs, modules)
4. Test user interactions with userEvent
5. Use descriptive test names
6. Keep tests focused and small
7. Avoid testing internal state
8. Test accessibility with role queries
9. Use story tests for visual testing
10. Run tests before committing