/sap-sac-test-automation

Playwright Execution Layer

Source: Derived summary from incorporated SAC automated-suite planning content and Playwright documentation.

Use Playwright as the deterministic execution layer for reviewed tests and CI. Load the local playwright skill for CLI browser-driving/debugging syntax when needed, but keep durable SAC regression suites in normal @playwright/test project code.

Configuration Defaults

Start conservative:

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

export default defineConfig({
  testDir: './tests',
  timeout: 120_000,
  expect: {
    timeout: 20_000,
    toHaveScreenshot: { maxDiffPixelRatio: 0.005 }
  },
  workers: process.env.CI ? 1 : undefined,
  retries: process.env.CI ? 2 : 0,
  reporter: [
    ['html', { outputFolder: 'playwright-report', open: 'never' }],
    ['junit', { outputFile: 'test-results/junit.xml' }],
    ['json', { outputFile: 'test-results/results.json' }]
  ],
  use: {
    baseURL: process.env.SAC_BASE_URL,
    viewport: { width: 1440, height: 1000 },
    locale: 'en-US',
    timezoneId: 'UTC',
    actionTimeout: 30_000,
    navigationTimeout: 90_000,
    screenshot: 'only-on-failure',
    video: 'retain-on-failure',
    trace: 'on-first-retry'
  },
  projects: [
    { name: 'chromium-linux', use: { ...devices['Desktop Chrome'] } }
  ]
});

Adjust to the project's existing package manager, CI provider, and artifact conventions.

Microsoft Edge Projects

Use installed Microsoft Edge when enterprise policy, user acceptance criteria, media behavior, extensions, certificates, or SSO require Edge:

projects: [
  { name: 'chromium-linux', use: { ...devices['Desktop Chrome'] } },
  { name: 'msedge', use: { ...devices['Desktop Edge'], channel: 'msedge' } }
]

Treat Edge channels as controlled-runner dependencies. Playwright can target msedge, msedge-beta, msedge-dev, and msedge-canary, but it does not install branded Edge by default in normal project setup. In companies with browser policies, mandatory extensions, or locked-down profiles, local Edge control may differ from CI behavior. Use bundled Chromium for local developer smoke work when policies interfere, and run Edge-channel regression on approved bots.

Use chromium.connectOverCDP('http://127.0.0.1:9222') only for approved local discovery or SSO/certificate triage against a running Edge instance. Do not build the main regression suite around CDP attach.

Restricted Installations

• Prefer pinned dependencies and the project's existing package manager.
• If public npm is blocked, use the organization's internal registry or approved package cache.
• If Playwright browser downloads are blocked, use an approved artifact mirror through PLAYWRIGHT_DOWNLOAD_HOST, set PLAYWRIGHT_BROWSERS_PATH to a pre-provisioned cache, and configure HTTPS_PROXY or NODE_EXTRA_CA_CERTS when required by corporate network policy.
• If no CLI tooling can be installed, produce a manual observation packet and keep deterministic Playwright implementation for an approved build agent.

Auth Strategy

• Use dedicated SAC test users.
• Store Playwright storageState securely outside Git.
• Maintain one auth state per role: viewer, planner, admin, restricted, no-export, or other project roles.
• Run an auth-check stage that opens SAC and fails fast if redirected to login.
• Avoid automating MFA as a release dependency. Use security-approved test policy or manual refresh when needed.

Readiness Strategy

Do not rely on page.goto() alone. Wait for:

• Story title or stable page marker.
• Critical widgets listed in the profile.
• Absence of configured SAC error text.
• Disappearance of busy/loading indicators.
• Stable table/KPI/chart state after interactions.
• Download, toast, prompt, or diagnostic/test-mode status where applicable.

CI Stages

Use staged execution:

install
lint-profiles
auth-check
read-only-smoke
nightly-regression
controlled-writeback
publish-report

Recommended gate policy:

• PR/push gate: read-only smoke, auth check, basic navigation, critical widget readiness.
• Nightly: filters, input controls, prompts, tables, visual baselines, data checks, exports, permissions.
• Controlled/nightly only: comments, planning, data actions, multi actions.
• Production: read-only smoke only unless formal approval says otherwise.

Test Category Notes

• Tables: locate by title/profile metadata, map headers, scroll/search virtualized rows, parse values by locale, compare with tolerances.
• Charts: use screenshots for rendering/layout and companion table/KPI assertions for business values.
• Filters/prompts: select by visible label/search, assert selected state and downstream data effect, reset after test.
• Scripts/custom buttons: test through UI-visible effects, diagnostic widgets, downloads, or data changes.
• Exports: wait for download, validate file type/name/size, parse content only when stable and valuable.
• Permissions: assert both UI visibility and data restrictions. Do not test only with admin users.
• Planning/comments: isolate, run serially, reset before/after, and use unique run IDs.