playwright

Playwright Test Writing

Overview

Write browser tests using @playwright/test, the official Playwright test runner. It provides auto-waiting, test isolation, built-in web assertions with auto-retry, and parallel execution out of the box.

Prefer @playwright/test for all testing scenarios. Use library mode (playwright) only for direct browser scripting needs: console error capture, network inspection, or custom automation outside a test context.

Project Setup

# Initialize Playwright in a project
npm init playwright@latest

# Or add to existing project
npm install -D @playwright/test
npx playwright install

Essential playwright.config.ts (for full multi-browser, reporter, and webServer config, see references/api-patterns.md):

import { defineConfig } from '@playwright/test';

export default defineConfig({
  testDir: './tests',
  fullyParallel: true,
  retries: process.env.CI ? 2 : 0,
  use: {
    baseURL: 'http://localhost:3000',
    trace: 'on-first-retry',
  },
});

Test File Structure

Tests follow the pattern: tests/<feature>.spec.ts

import { test, expect } from '@playwright/test';

test.describe('Feature Name', () => {
  test.beforeEach(async ({ page }) => {
    await page.goto('/');
  });

  test('should perform expected behavior', async ({ page }) => {
    // Arrange
    const submitButton = page.getByRole('button', { name: 'Submit' });

    // Act
    await submitButton.click();

    // Assert
    await expect(page.getByText('Success')).toBeVisible();
  });
});

Always use the AAA pattern: Arrange (locate elements), Act (interact), Assert (verify outcome).

Locator Strategy (Precision Order)

Apply locators in strict priority order — always prefer the highest-precision option available:

Priority	Locator	When to Use	Stability
1	getByTestId('id')	Element has data-testid attribute	Highest — immune to text/structure changes
2	getByRole('role', { name })	Element has clear ARIA role and accessible name	Very high — maps from accessibility tree
3	getByLabel('label')	Form inputs with associated <label>	High — tied to label text
4	getByPlaceholder('text')	Input with placeholder text	Medium-high — placeholder may change
5	getByText('text', { exact: true })	Visible text content	Medium — text may change
6	page.locator('[data-attr="val"]')	Custom data attributes	Medium — depends on attribute stability
7	page.locator('css')	Last resort — specific CSS selector	Low — fragile to DOM restructuring

Rules:

• Never use index-based locators (nth(0)) unless testing a list where index is semantically meaningful.
• Never use XPath unless no other option exists.
• Always pass { exact: true } to getByText() to prevent partial matches.
• Prefer accessibility-based locators (getByRole, getByLabel) — they match what users see and interact with.
• When multiple elements share the same locator, narrow with .filter({ hasText: 'unique' }) or scope to a parent.

For locator disambiguation patterns (filter, chaining, scoping), see references/api-patterns.md.

Web Assertions (Auto-retry)

Playwright's expect API auto-retries assertions until timeout (default 5s):

// Visibility
await expect(locator).toBeVisible();
await expect(locator).toBeHidden();

// Text content
await expect(locator).toHaveText('exact text');
await expect(locator).toContainText('partial');

// Input values
await expect(locator).toHaveValue('input value');

// Attributes
await expect(locator).toHaveAttribute('href', '/path');

// Count
await expect(locator).toHaveCount(3);

// Element state
await expect(locator).toBeEnabled();
await expect(locator).toBeDisabled();
await expect(locator).toBeChecked();

// Page-level
await expect(page).toHaveURL(/\/dashboard/);
await expect(page).toHaveTitle('Dashboard');

Critical: Always use await expect(locator).toHaveText() — not expect(await locator.textContent()).toBe(). Web assertions auto-retry; manual extraction does not.

Page Interactions

// Click
await page.getByRole('button', { name: 'Save' }).click();

// Fill (clears existing content first)
await page.getByLabel('Email').fill('[email protected]');

// Type sequentially (appends, triggers key events)
await page.getByLabel('Search').pressSequentially('query');

// Select dropdown
await page.getByLabel('Country').selectOption('US');

// Check / uncheck
await page.getByLabel('Agree to terms').check();

// Keyboard
await page.keyboard.press('Enter');
await page.keyboard.press('Escape');

// File upload
await page.getByLabel('Upload').setInputFiles('path/to/file.pdf');

// Wait for navigation
await page.waitForURL('/dashboard');

Test Lifecycle

test.describe('Suite', () => {
  test.beforeAll(async () => {
    // Run once before all tests in this suite
  });

  test.beforeEach(async ({ page }) => {
    // Run before each test — common setup (e.g., navigation)
  });

  test.afterEach(async ({ page }) => {
    // Run after each test — cleanup
  });

  test.afterAll(async () => {
    // Run once after all tests in this suite
  });
});

Custom Fixtures

Extend the test object with reusable setup logic:

import { test as base, Page } from '@playwright/test';

const test = base.extend<{ authenticatedPage: Page }>({
  authenticatedPage: async ({ page }, use) => {
    await page.goto('/login');
    await page.getByLabel('Email').fill('[email protected]');
    await page.getByLabel('Password').fill('password');
    await page.getByRole('button', { name: 'Sign In' }).click();
    await page.waitForURL('/dashboard');
    await use(page);
  },
});

test('should show user profile', async ({ authenticatedPage }) => {
  await authenticatedPage.goto('/profile');
  await expect(authenticatedPage.getByRole('heading', { name: 'Profile' })).toBeVisible();
});

Running Tests

npx playwright test                    # Run all tests
npx playwright test login.spec.ts      # Run specific file
npx playwright test --headed           # Visible browser
npx playwright test --ui               # Interactive UI mode
npx playwright test --debug            # Step-through debugger
npx playwright show-report             # View HTML report

Library Mode (Advanced)

Use playwright (not @playwright/test) when direct browser control is needed outside a test runner:

import { chromium } from 'playwright';

const browser = await chromium.launch();
const context = await browser.newContext();
const page = await context.newPage();

// Console error capture
page.on('console', msg => {
  if (msg.type() === 'error') console.log('Console error:', msg.text());
});

// Network request inspection
page.on('response', response => {
  if (response.status() >= 400)
    console.log(`Failed: ${response.status()} ${response.url()}`);
});

// JavaScript execution in page context
const title = await page.evaluate(() => document.title);

// Cookie inspection
const cookies = await context.cookies();

await browser.close();

Use library mode for: console error capture, network inspection, page.evaluate(), iframe/shadow DOM exploration, cookie/localStorage checks.

Additional Resources

Reference Files

For detailed locator mapping, assertion catalog, and advanced configuration:

• references/api-patterns.md — Locator disambiguation patterns, assertion quick reference, advanced fixtures and configuration