webship-js 2.0.11 → 2.0.13
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.circleci/config.yml +10 -1
- package/.drone.yml +7 -7
- package/.github/workflows/github-actions.yml +48 -0
- package/.gitlab-ci.yml +9 -4
- package/.harness/webship-js-pipeline.yml +5 -6
- package/.semaphore/semaphore.yml +6 -9
- package/.teamcity/settings.kts +7 -6
- package/.travis.yml +6 -1
- package/AGENTS.md +1 -1
- package/CLAUDE.md +65 -23
- package/Jenkinsfile +5 -3
- package/azure-pipelines.yml +17 -8
- package/bamboo-specs/bamboo.yml +9 -15
- package/bin/init-webship.js +53 -10
- package/bitbucket-pipelines.yml +5 -1
- package/buildspec.yml +6 -7
- package/cloudbuild.yaml +28 -12
- package/codefresh.yml +6 -6
- package/cucumber.js +17 -10
- package/docs/00-quick-start.md +8 -0
- package/docs/02-bbr-smart-waits.md +15 -10
- package/docs/03-selector-registry.md +24 -0
- package/docs/04-step-reference.md +51 -25
- package/docs/05-web-first-assertions.md +6 -0
- package/docs/07-auth-state.md +1 -1
- package/docs/08-clock-mocking.md +21 -2
- package/docs/09-api-testing.md +40 -2
- package/docs/10-accessibility.md +2 -0
- package/docs/11-debugging.md +3 -1
- package/docs/12-ai-agent-guide.md +5 -0
- package/docs/13-faq.md +72 -1
- package/docs/14-recipes-cookbook.md +2 -0
- package/docs/15-tag-conventions.md +2 -1
- package/docs/16-ci-cd.md +140 -119
- package/docs/README.md +5 -5
- package/docs/advanced-screenshots.md +3 -1
- package/docs/api-step-definitions.md +16 -2
- package/docs/global-settings.md +69 -17
- package/docs/install-webship-js.md +2 -0
- package/docs/step-definitions/then-i-should-see.md +14 -1
- package/docs/step-definitions/then-item-checkbox.md +2 -0
- package/docs/step-definitions/when-i-fill.md +4 -0
- package/docs/step-definitions/when-i-select-option-select-list.md +2 -0
- package/examples/check-switch.html +37 -0
- package/examples/choice-controls.html +29 -0
- package/examples/click-whitespace.html +37 -0
- package/examples/datetime-fields.html +52 -0
- package/examples/element-selector-hint.html +16 -0
- package/examples/field-states.html +27 -0
- package/examples/fill-in-empty.html +26 -0
- package/examples/fill-unreachable.html +27 -0
- package/examples/follow-hidden-first.html +29 -0
- package/examples/iframe-inner.html +20 -0
- package/examples/iframe-nested.html +16 -0
- package/examples/iframe-outer.html +15 -0
- package/examples/multi-value-fields.html +54 -0
- package/examples/registered-names.html +15 -0
- package/examples/select-by-label.html +33 -0
- package/examples/settle-budget.html +21 -0
- package/examples/sorted-tables.html +38 -0
- package/examples/text-matching.html +48 -0
- package/examples/text-order.html +25 -0
- package/examples/touch.html +22 -0
- package/examples/viewport-top.html +36 -0
- package/package.json +17 -1
- package/playwright.config.ts +3 -1
- package/tests/check-api-launch-url.js +106 -0
- package/tests/check-assets.js +103 -0
- package/tests/check-attachments.js +105 -0
- package/tests/check-breakpoints.js +117 -0
- package/tests/check-clock.js +143 -0
- package/tests/check-config.js +90 -0
- package/tests/check-context.js +215 -0
- package/tests/check-dry-run.js +104 -0
- package/tests/check-examples.js +10 -4
- package/tests/check-exit-report.js +141 -0
- package/tests/check-javascript-errors.js +105 -0
- package/tests/check-must-fail.js +5 -4
- package/tests/check-network-changed.js +137 -0
- package/tests/check-scaffold.js +54 -2
- package/tests/check-screenshots.js +117 -0
- package/tests/check-settle-verbs.js +137 -0
- package/tests/check-time-limit.js +104 -0
- package/tests/check-video.js +156 -0
- package/tests/check-world.js +167 -0
- package/tests/features/api-base-url.feature +25 -0
- package/tests/features/api-form-data.feature +18 -0
- package/tests/features/api-json-paths.feature +24 -0
- package/tests/features/api-response-contains.feature +21 -0
- package/tests/features/api-response-header-presence.feature +17 -0
- package/tests/features/api-response-header.feature +17 -0
- package/tests/features/api-saved-placeholders.feature +37 -0
- package/tests/features/api-step-definitions-individual.feature +1 -1
- package/tests/features/api-tables.feature +40 -0
- package/tests/features/breakpoints.feature +25 -0
- package/tests/features/check-switch.feature +28 -0
- package/tests/features/choice-controls.feature +35 -0
- package/tests/features/click-whitespace.feature +29 -0
- package/tests/features/complete-api-testing-examples.feature +1 -1
- package/tests/features/datetime-field-lookup.feature +29 -0
- package/tests/features/element-selector-hint.feature +14 -0
- package/tests/features/field-state.feature +21 -0
- package/tests/features/fill-in-empty-value.feature +41 -0
- package/tests/features/fill-reachable.feature +13 -0
- package/tests/features/follow-hidden-first.feature +9 -0
- package/tests/features/iframe-nested.feature +43 -0
- package/tests/features/javascript-tolerated-errors.feature +20 -0
- package/tests/features/modal-open.feature +13 -0
- package/tests/features/multi-value-field-lookup.feature +30 -0
- package/tests/features/registered-names.feature +14 -0
- package/tests/features/select-by-label.feature +31 -0
- package/tests/features/selector-files.feature +14 -0
- package/tests/features/selectors-table-header.feature +35 -0
- package/tests/features/settle-budget.feature +11 -0
- package/tests/features/sorted-tables.feature +19 -0
- package/tests/features/storage-quoted-values.feature +26 -0
- package/tests/features/text-matching.feature +38 -0
- package/tests/features/text-order.feature +14 -0
- package/tests/features/touch.feature +21 -0
- package/tests/features/viewport-top.feature +13 -0
- package/tests/features/xml-namespaces.feature +34 -0
- package/tests/must-fail/api-json-paths.feature +21 -0
- package/tests/must-fail/api-response-contains.feature +23 -0
- package/tests/must-fail/api-response-header-presence.feature +17 -0
- package/tests/must-fail/api-response-header.feature +21 -0
- package/tests/must-fail/api-without-base-url.feature +11 -0
- package/tests/must-fail/breakpoints.feature +15 -0
- package/tests/must-fail/choice-controls.feature +25 -0
- package/tests/must-fail/clock.feature +11 -0
- package/tests/must-fail/datetime-field.feature +10 -0
- package/tests/must-fail/element-selector-hint.feature +17 -0
- package/tests/must-fail/field-state.feature +20 -0
- package/tests/must-fail/fill-field.feature +30 -0
- package/tests/must-fail/iframe-text.feature +17 -0
- package/tests/must-fail/modal-open.feature +26 -0
- package/tests/must-fail/multi-value-field.feature +18 -0
- package/tests/must-fail/registered-names.feature +25 -0
- package/tests/must-fail/select-list.feature +15 -0
- package/tests/must-fail/selector-files.feature +10 -0
- package/tests/must-fail/sorted-tables.feature +20 -0
- package/tests/must-fail/storage-before-a-page.feature +13 -0
- package/tests/must-fail/text-matching.feature +16 -0
- package/tests/must-fail/text-order.feature +15 -0
- package/tests/must-fail/viewport-top.feature +20 -0
- package/tests/must-fail/yaml-numbers.feature +48 -0
- package/tests/step-definitions/a11y.steps.js +24 -15
- package/tests/step-definitions/action.steps.js +53 -20
- package/tests/step-definitions/api.steps.js +349 -67
- package/tests/step-definitions/assertion.steps.js +100 -10
- package/tests/step-definitions/auth.steps.js +15 -18
- package/tests/step-definitions/clock.steps.js +22 -0
- package/tests/step-definitions/debug.steps.js +8 -0
- package/tests/step-definitions/dialog.steps.js +7 -0
- package/tests/step-definitions/element.steps.js +132 -17
- package/tests/step-definitions/field.steps.js +412 -96
- package/tests/step-definitions/form.steps.js +88 -37
- package/tests/step-definitions/iframe.steps.js +89 -29
- package/tests/step-definitions/input.steps.js +26 -7
- package/tests/step-definitions/javascript.steps.js +64 -22
- package/tests/step-definitions/modal.steps.js +79 -13
- package/tests/step-definitions/navigation.steps.js +14 -0
- package/tests/step-definitions/network.steps.js +15 -5
- package/tests/step-definitions/response.steps.js +6 -4
- package/tests/step-definitions/responsive.steps.js +26 -33
- package/tests/step-definitions/screenshot.steps.js +25 -8
- package/tests/step-definitions/scroll.steps.js +20 -0
- package/tests/step-definitions/selectors.steps.js +95 -48
- package/tests/step-definitions/storage.steps.js +38 -13
- package/tests/step-definitions/table.steps.js +81 -6
- package/tests/step-definitions/video.steps.js +52 -55
- package/tests/step-definitions/wait.steps.js +88 -23
- package/tests/step-definitions/web-first.steps.js +69 -35
- package/tests/step-definitions/webship.js +741 -145
- package/tests/step-definitions/xml.steps.js +49 -15
- package/tests/step-definitions/yaml.steps.js +45 -20
package/bin/init-webship.js
CHANGED
|
@@ -72,7 +72,10 @@ module.exports = {
|
|
|
72
72
|
xpath: {},
|
|
73
73
|
filesPath: './tests/selectors/',
|
|
74
74
|
files: [],
|
|
75
|
-
|
|
75
|
+
// Named viewport sizes for the viewing-the-site and set-the-viewport
|
|
76
|
+
// steps and the @mobile / @desktop tags (built in: xs..xxxl, mobile,
|
|
77
|
+
// tablet, desktop). Without a desktop here, @desktop opens at the
|
|
78
|
+
// breakpoint marked default: true.
|
|
76
79
|
breakpoints: {
|
|
77
80
|
xs: { width: 375, height: 667 },
|
|
78
81
|
sm: { width: 576, height: 800 },
|
|
@@ -126,7 +129,9 @@ const config: PlaywrightConfig = {
|
|
|
126
129
|
browser,
|
|
127
130
|
launchOptions: {
|
|
128
131
|
headless: process.env.HEADLESS !== 'false',
|
|
129
|
-
//
|
|
132
|
+
// Pause before every browser action, in ms: 300 by default, 800 with
|
|
133
|
+
// HEADLESS=false so a headed run can be watched. SLOW_MO wins over both:
|
|
134
|
+
// SLOW_MO=0 for full speed (npm run test:fast), in CI too.
|
|
130
135
|
slowMo: parseInt(process.env.SLOW_MO || (process.env.HEADLESS === 'false' ? '800' : '300'), 10),
|
|
131
136
|
args: browser === 'chromium' ? chromiumArgs : [],
|
|
132
137
|
},
|
|
@@ -192,6 +197,15 @@ const TEST_SCRIPTS = {
|
|
|
192
197
|
// the project, so the scaffold replaces it.
|
|
193
198
|
const NPM_PLACEHOLDER_TEST = /^echo "Error: no test specified" && exit 1$/;
|
|
194
199
|
|
|
200
|
+
// A `test` script that runs Nightwatch is the one webship-js 1.x gave a
|
|
201
|
+
// project. webship-js 2.x runs on cucumber-js and Playwright and brings no
|
|
202
|
+
// Nightwatch, so the scaffold replaces it and keeps it as `test:nightwatch`.
|
|
203
|
+
const NIGHTWATCH_TEST = /(^|[\s;&|/])nightwatch(\s|$)/i;
|
|
204
|
+
|
|
205
|
+
// Files of the scaffold that hold the project's own code once written: never
|
|
206
|
+
// overwritten, not even with --force.
|
|
207
|
+
const PROJECT_FILES = new Set(['tests/step-definitions/custom.js']);
|
|
208
|
+
|
|
195
209
|
const DEFAULT_PKG = {
|
|
196
210
|
name: 'project-name',
|
|
197
211
|
version: '2.0.0',
|
|
@@ -227,7 +241,8 @@ function printHelp() {
|
|
|
227
241
|
'Idempotent — safe to run multiple times.',
|
|
228
242
|
'',
|
|
229
243
|
'Options:',
|
|
230
|
-
' -f, --force Overwrite
|
|
244
|
+
' -f, --force Overwrite the files and scripts of the scaffold',
|
|
245
|
+
' (never tests/step-definitions/custom.js)',
|
|
231
246
|
' --skip-browsers Skip `npx playwright install chromium`',
|
|
232
247
|
' --cwd <dir> Target directory (default: $INIT_CWD or cwd)',
|
|
233
248
|
' -h, --help Show this help',
|
|
@@ -263,7 +278,8 @@ function writeTextAssets(projectPath, opts) {
|
|
|
263
278
|
const results = [];
|
|
264
279
|
for (const [rel, content] of Object.entries(TEXT_ASSETS)) {
|
|
265
280
|
const dest = path.join(projectPath, rel);
|
|
266
|
-
const
|
|
281
|
+
const force = opts.force && !PROJECT_FILES.has(rel);
|
|
282
|
+
const written = writeFileIfMissing(dest, content, { force });
|
|
267
283
|
results.push({ file: rel, written });
|
|
268
284
|
}
|
|
269
285
|
return results;
|
|
@@ -288,20 +304,34 @@ function ensurePackageJson(projectPath, opts) {
|
|
|
288
304
|
}
|
|
289
305
|
|
|
290
306
|
pkg.scripts = pkg.scripts || {};
|
|
307
|
+
let replacedNightwatch = null;
|
|
291
308
|
for (const [name, cmd] of Object.entries(TEST_SCRIPTS)) {
|
|
292
|
-
const
|
|
293
|
-
|
|
309
|
+
const current = String(pkg.scripts[name] || '').trim();
|
|
310
|
+
const placeholder = name === 'test' && NPM_PLACEHOLDER_TEST.test(current);
|
|
311
|
+
const nightwatch = name === 'test' && NIGHTWATCH_TEST.test(current);
|
|
312
|
+
if (nightwatch && current !== cmd) {
|
|
313
|
+
if (!pkg.scripts['test:nightwatch']) pkg.scripts['test:nightwatch'] = pkg.scripts.test;
|
|
314
|
+
replacedNightwatch = pkg.scripts.test;
|
|
315
|
+
}
|
|
316
|
+
if (opts.force || !pkg.scripts[name] || placeholder || nightwatch) {
|
|
294
317
|
pkg.scripts[name] = cmd;
|
|
295
318
|
}
|
|
296
319
|
}
|
|
297
320
|
|
|
298
|
-
|
|
299
|
-
|
|
321
|
+
// webship-js is listed once: where the project already declares it (a dev or peer dependency
|
|
322
|
+
// included), with its range; --force only refreshes that range. Otherwise it goes under
|
|
323
|
+
// dependencies.
|
|
324
|
+
const declaredIn = ['dependencies', 'devDependencies', 'peerDependencies', 'optionalDependencies']
|
|
325
|
+
.find((field) => pkg[field] && pkg[field]['webship-js']);
|
|
326
|
+
if (!declaredIn) {
|
|
327
|
+
pkg.dependencies = pkg.dependencies || {};
|
|
300
328
|
pkg.dependencies['webship-js'] = readPkgVersion();
|
|
329
|
+
} else if (opts.force) {
|
|
330
|
+
pkg[declaredIn]['webship-js'] = readPkgVersion();
|
|
301
331
|
}
|
|
302
332
|
|
|
303
333
|
fs.writeFileSync(pkgPath, JSON.stringify(pkg, null, 2) + '\n', 'utf8');
|
|
304
|
-
return { created: !existed };
|
|
334
|
+
return { created: !existed, replacedNightwatch };
|
|
305
335
|
}
|
|
306
336
|
|
|
307
337
|
// The browser itself goes into the Playwright cache of the user: no special
|
|
@@ -352,12 +382,25 @@ function run(argv) {
|
|
|
352
382
|
const pkgResult = ensurePackageJson(projectPath, opts);
|
|
353
383
|
|
|
354
384
|
for (const r of assetResults) {
|
|
355
|
-
|
|
385
|
+
const note = !r.written && opts.force && PROJECT_FILES.has(r.file)
|
|
386
|
+
? ' (the steps of the project: --force never overwrites it)'
|
|
387
|
+
: '';
|
|
388
|
+
console.log(` [${r.written ? 'wrote ' : 'kept '}] ${r.file}${note}`);
|
|
356
389
|
}
|
|
357
390
|
console.log(` [${tsconfigWritten ? 'wrote ' : 'kept '}] tsconfig.json`);
|
|
358
391
|
console.log(
|
|
359
392
|
` [${pkgResult.created ? 'wrote ' : 'merged'}] package.json`
|
|
360
393
|
);
|
|
394
|
+
if (pkgResult.replacedNightwatch) {
|
|
395
|
+
console.log(
|
|
396
|
+
[
|
|
397
|
+
'',
|
|
398
|
+
' [webship-js] "npm test" ran Nightwatch (a webship-js 1.x project). It now runs',
|
|
399
|
+
' cucumber-js: cucumber-js --config cucumber.js',
|
|
400
|
+
' The old command is kept as "npm run test:nightwatch".',
|
|
401
|
+
].join('\n')
|
|
402
|
+
);
|
|
403
|
+
}
|
|
361
404
|
|
|
362
405
|
const browsersInstalled = opts.skipBrowsers || installBrowsers(projectPath);
|
|
363
406
|
|
package/bitbucket-pipelines.yml
CHANGED
package/buildspec.yml
CHANGED
|
@@ -2,11 +2,11 @@
|
|
|
2
2
|
#
|
|
3
3
|
# Runs the full BDD suite on AWS-managed compute. Pair with CodePipeline
|
|
4
4
|
# (for source/build/test orchestration) or invoke CodeBuild directly from a
|
|
5
|
-
# GitHub webhook.
|
|
6
|
-
#
|
|
5
|
+
# GitHub webhook. For the free tier and the prices, see the CodeBuild pricing
|
|
6
|
+
# page: they change over time.
|
|
7
7
|
#
|
|
8
8
|
# Suggested CodeBuild project settings:
|
|
9
|
-
# Environment image: aws/codebuild/standard:
|
|
9
|
+
# Environment image: aws/codebuild/standard:8.0 (Ubuntu 24.04, Node 22/24/26)
|
|
10
10
|
# Compute: general1.small (3 GB RAM, 2 vCPU) — fits the suite
|
|
11
11
|
# Service role: needs S3 PutObject for the artifacts bucket
|
|
12
12
|
#
|
|
@@ -23,7 +23,7 @@ env:
|
|
|
23
23
|
phases:
|
|
24
24
|
install:
|
|
25
25
|
runtime-versions:
|
|
26
|
-
nodejs:
|
|
26
|
+
nodejs: 24
|
|
27
27
|
commands:
|
|
28
28
|
- node --version
|
|
29
29
|
- npm --version
|
|
@@ -49,8 +49,7 @@ cache:
|
|
|
49
49
|
artifacts:
|
|
50
50
|
name: webship-js-report-$(date +%Y%m%d-%H%M%S)
|
|
51
51
|
files:
|
|
52
|
-
- tests/reports
|
|
53
|
-
- tests/reports/cucumber_report.pdf
|
|
54
|
-
- tests/reports/cucumber_report.json
|
|
52
|
+
- tests/reports/**/*
|
|
55
53
|
- screenshots/**/*
|
|
54
|
+
- videos/**/*
|
|
56
55
|
discard-paths: no
|
package/cloudbuild.yaml
CHANGED
|
@@ -1,9 +1,10 @@
|
|
|
1
1
|
# Google Cloud Build config for webship-js.
|
|
2
2
|
#
|
|
3
3
|
# Runs the full BDD suite inside a Playwright-prepped container on Google's
|
|
4
|
-
# managed builder pool.
|
|
5
|
-
#
|
|
6
|
-
#
|
|
4
|
+
# managed builder pool. For the free tier and the prices, see the Cloud Build
|
|
5
|
+
# pricing page: they change over time. The image is the one of the Playwright
|
|
6
|
+
# the install brings; `npx playwright install` gets the browser of a newer
|
|
7
|
+
# Playwright release when there is one.
|
|
7
8
|
#
|
|
8
9
|
# Setup (full walkthrough in docs/16-ci-cd.md → "Google Cloud Build"):
|
|
9
10
|
# 1. console.cloud.google.com → create project → enable Cloud Build API.
|
|
@@ -22,7 +23,7 @@ timeout: '1800s'
|
|
|
22
23
|
|
|
23
24
|
steps:
|
|
24
25
|
- id: install-and-test
|
|
25
|
-
name: 'mcr.microsoft.com/playwright:v1.
|
|
26
|
+
name: 'mcr.microsoft.com/playwright:v1.65.1-noble'
|
|
26
27
|
entrypoint: 'bash'
|
|
27
28
|
env:
|
|
28
29
|
- 'FORCE_COLOR=1'
|
|
@@ -38,12 +39,27 @@ steps:
|
|
|
38
39
|
nohup npm start > /tmp/srv.log 2>&1 &
|
|
39
40
|
sleep 3
|
|
40
41
|
curl -sf http://localhost:8080/ > /dev/null
|
|
41
|
-
|
|
42
|
+
# A failing suite must not stop the build before the report is kept:
|
|
43
|
+
# the result is checked by the last step.
|
|
44
|
+
npm test || touch /workspace/.tests-failed
|
|
42
45
|
|
|
43
|
-
artifacts
|
|
44
|
-
|
|
45
|
-
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
|
|
49
|
-
|
|
46
|
+
# Cloud Build uploads the files of `artifacts:` only when every step
|
|
47
|
+
# succeeds, so the report, the screenshots and the videos are copied here,
|
|
48
|
+
# passed or failed.
|
|
49
|
+
- id: keep-report
|
|
50
|
+
name: 'gcr.io/cloud-builders/gsutil'
|
|
51
|
+
entrypoint: 'bash'
|
|
52
|
+
args:
|
|
53
|
+
- '-c'
|
|
54
|
+
- |
|
|
55
|
+
for dir in tests/reports screenshots videos; do
|
|
56
|
+
if [ -d "$$dir" ]; then gsutil -m cp -r "$$dir" "gs://${_ARTIFACT_BUCKET}/$BUILD_ID/"; fi
|
|
57
|
+
done
|
|
58
|
+
|
|
59
|
+
- id: test-result
|
|
60
|
+
name: 'gcr.io/cloud-builders/gsutil'
|
|
61
|
+
entrypoint: 'bash'
|
|
62
|
+
args:
|
|
63
|
+
- '-c'
|
|
64
|
+
- |
|
|
65
|
+
if [ -f /workspace/.tests-failed ]; then echo "The suite failed: see the report in gs://${_ARTIFACT_BUCKET}/$BUILD_ID/"; exit 1; fi
|
package/codefresh.yml
CHANGED
|
@@ -1,9 +1,8 @@
|
|
|
1
1
|
# Codefresh Classic pipeline for webship-js.
|
|
2
2
|
#
|
|
3
3
|
# Codefresh is a Kubernetes-native CI/CD platform. The Classic pipeline format
|
|
4
|
-
# (used here) reads `codefresh.yml` from the repository.
|
|
5
|
-
# build
|
|
6
|
-
# can apply for the Community plan with more capacity.
|
|
4
|
+
# (used here) reads `codefresh.yml` from the repository. For the plans and
|
|
5
|
+
# their build capacity, see the Codefresh pricing page: they change over time.
|
|
7
6
|
#
|
|
8
7
|
# Setup (full walkthrough in docs/16-ci-cd.md → "Codefresh"):
|
|
9
8
|
# 1. codefresh.io → sign up with GitHub → create pipeline → "Inline YAML
|
|
@@ -34,7 +33,7 @@ steps:
|
|
|
34
33
|
install_dependencies:
|
|
35
34
|
title: "Install npm dependencies + Playwright browsers"
|
|
36
35
|
type: freestyle
|
|
37
|
-
image: mcr.microsoft.com/playwright:v1.
|
|
36
|
+
image: mcr.microsoft.com/playwright:v1.65.1-noble
|
|
38
37
|
working_directory: ${{main_clone}}
|
|
39
38
|
environment:
|
|
40
39
|
- FORCE_COLOR=1
|
|
@@ -48,7 +47,7 @@ steps:
|
|
|
48
47
|
test_suite:
|
|
49
48
|
title: "Run cucumber-js"
|
|
50
49
|
type: freestyle
|
|
51
|
-
image: mcr.microsoft.com/playwright:v1.
|
|
50
|
+
image: mcr.microsoft.com/playwright:v1.65.1-noble
|
|
52
51
|
working_directory: ${{main_clone}}
|
|
53
52
|
environment:
|
|
54
53
|
- FORCE_COLOR=1
|
|
@@ -72,6 +71,7 @@ steps:
|
|
|
72
71
|
failureToo: "true"
|
|
73
72
|
commands:
|
|
74
73
|
- mkdir -p /codefresh/volume/cucumber-report
|
|
75
|
-
- cp -a tests/reports
|
|
74
|
+
- cp -a tests/reports/. /codefresh/volume/cucumber-report/ 2>/dev/null || true
|
|
76
75
|
- cp -a screenshots /codefresh/volume/cucumber-report/screenshots 2>/dev/null || true
|
|
76
|
+
- cp -a videos /codefresh/volume/cucumber-report/videos 2>/dev/null || true
|
|
77
77
|
stage: test
|
package/cucumber.js
CHANGED
|
@@ -91,9 +91,12 @@ module.exports = {
|
|
|
91
91
|
// Load additional selectors from JSON files at scenario start.
|
|
92
92
|
filesPath: './tests/selectors/', // base path for selector files
|
|
93
93
|
files: [], // e.g. ['homepage-selectors.json'] or ['front-end-selectors.json', 'back-end-selectors.json']
|
|
94
|
-
//
|
|
95
|
-
|
|
96
|
-
//
|
|
94
|
+
// Named viewport sizes for every step and tag that names one (viewing
|
|
95
|
+
// the site on a … screen, set the viewport to the "…" breakpoint,
|
|
96
|
+
// @mobile, @desktop). Built in: xs..xxxl, mobile, tablet, desktop;
|
|
97
|
+
// these replace or add names. WEBSHIP_SELECTORS_BREAKPOINTS (a JSON
|
|
98
|
+
// object) is merged over them. Without a `desktop` here, @desktop
|
|
99
|
+
// opens at the breakpoint marked default: true.
|
|
97
100
|
breakpoints: {
|
|
98
101
|
xs: { width: 375, height: 667 }, // phone portrait
|
|
99
102
|
sm: { width: 576, height: 800 }, // large phone / phablet
|
|
@@ -117,19 +120,23 @@ module.exports = {
|
|
|
117
120
|
},
|
|
118
121
|
video: {
|
|
119
122
|
// Record the browser as a webm video.
|
|
120
|
-
// 'off' (default) — no recording.
|
|
123
|
+
// 'off' (default) — no recording, except scenarios tagged @video.
|
|
121
124
|
// 'on' — record every scenario.
|
|
122
|
-
// 'on-failure' — record every scenario, keep only failures
|
|
123
|
-
//
|
|
125
|
+
// 'on-failure' — record every scenario, keep only failures (and
|
|
126
|
+
// the videos of @video scenarios or with a name
|
|
127
|
+
// reserved by "I save the current video as").
|
|
128
|
+
// 'tag' — the same as 'off': only scenarios tagged @video.
|
|
124
129
|
// Override per scenario with tags:
|
|
125
|
-
// @video — force recording on (
|
|
130
|
+
// @video — force recording on (in every mode).
|
|
126
131
|
// @no-video — suppress recording for this scenario.
|
|
127
132
|
// Override per run with env: WEBSHIP_VIDEO.
|
|
128
133
|
mode: 'off', // WEBSHIP_VIDEO
|
|
129
134
|
dir: './videos', // WEBSHIP_VIDEO_DIR
|
|
130
|
-
size: { width: 1280, height: 720 }, //
|
|
131
|
-
// Filename template tokens: {datetime}
|
|
132
|
-
// {
|
|
135
|
+
size: { width: 1280, height: 720 }, // WEBSHIP_VIDEO_SIZE e.g. "1280x720"
|
|
136
|
+
// Filename template tokens: {datetime} (when the recording started)
|
|
137
|
+
// {feature_file} {feature} {scenario} {status} (passed, failed, … in
|
|
138
|
+
// lower case; "manual" for a recording stopped by a step) {ext}.
|
|
139
|
+
// Sanitised to filesystem-safe characters.
|
|
133
140
|
filenamePattern: '{datetime}.{feature_file}.{scenario}.{status}.{ext}',
|
|
134
141
|
},
|
|
135
142
|
javascript: {
|
package/docs/00-quick-start.md
CHANGED
|
@@ -55,6 +55,14 @@ npx init-webship-js
|
|
|
55
55
|
npm install-scripts approve webship-js
|
|
56
56
|
```
|
|
57
57
|
|
|
58
|
+
The scaffold never replaces a file or a `test` script the project already
|
|
59
|
+
has, with two exceptions: npm's "no test specified" placeholder, and a
|
|
60
|
+
`test` script that runs Nightwatch (a project upgraded from webship-js
|
|
61
|
+
1.x), which becomes the cucumber-js suite while the old command stays
|
|
62
|
+
available as `npm run test:nightwatch`. `npx init-webship-js --force`
|
|
63
|
+
rewrites the scaffold's files and scripts, but never
|
|
64
|
+
`tests/step-definitions/custom.js`, which holds the project's own steps.
|
|
65
|
+
|
|
58
66
|
Firefox and WebKit are optional: `npx playwright install firefox webkit`.
|
|
59
67
|
|
|
60
68
|
## 60-second first test
|
|
@@ -32,21 +32,23 @@ reads the count the same way.
|
|
|
32
32
|
### How the budget is shared
|
|
33
33
|
|
|
34
34
|
Each phase is best-effort — a phase that times out is not a failure — and each
|
|
35
|
-
gets a **share of the budget** rather than "whatever is left":
|
|
35
|
+
but the last gets a **share of the budget** rather than "whatever is left":
|
|
36
36
|
|
|
37
37
|
| Phase | Share of the budget |
|
|
38
38
|
| --- | --- |
|
|
39
39
|
| `<body>` attached | 25% |
|
|
40
40
|
| `DOMContentLoaded` | 25% |
|
|
41
41
|
| `networkidle` | 50% |
|
|
42
|
-
| AJAX + pending timers + DOM quiet |
|
|
42
|
+
| AJAX + pending timers + DOM quiet | everything left |
|
|
43
43
|
|
|
44
|
-
|
|
45
|
-
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
|
|
49
|
-
|
|
44
|
+
A phase that finishes early hands its unspent time to the ones after it, and
|
|
45
|
+
no phase is ever given more than what is actually left. Capping `networkidle`
|
|
46
|
+
matters, because on a page with analytics, a chat widget or a
|
|
47
|
+
self-rescheduling timer it never arrives — uncapped it would eat the whole
|
|
48
|
+
budget and starve the AJAX / pending-timer / DOM-quiet probe, which is the
|
|
49
|
+
phase that decides the answer. That probe comes last, so it takes the whole
|
|
50
|
+
rest of the budget: after a step that changes the page (1.5 seconds), a
|
|
51
|
+
request or a timer of a second is still waited for.
|
|
50
52
|
|
|
51
53
|
`smartSettle` returns within `budget` even when nothing on the page ever
|
|
52
54
|
settles, and never hands Playwright a zero timeout (Playwright reads `0` as
|
|
@@ -82,14 +84,17 @@ When I wait for the modal to disappear
|
|
|
82
84
|
# Body + DOMContentLoaded only (cheaper than full smart settle).
|
|
83
85
|
When I wait until the page is interactive
|
|
84
86
|
|
|
85
|
-
# Polling text assertion — re-runs the matcher until the deadline.
|
|
87
|
+
# Polling text assertion — re-runs the matcher until the deadline. It reads the
|
|
88
|
+
# page as "I should see" does: letter case ignored, hidden text included.
|
|
86
89
|
Then eventually I should see "Done"
|
|
87
90
|
Then eventually I should see "Done" within 10 seconds
|
|
88
91
|
```
|
|
89
92
|
|
|
90
93
|
## Auto-settle after actions
|
|
91
94
|
|
|
92
|
-
After every state-changing step
|
|
95
|
+
After every state-changing step, webship-js silently runs `smartSettle(page, 1500)`. Tests do not need an explicit wait between an action and its follow-up assertion in the typical case. Disable per-run with `WEBSHIP_AUTO_SETTLE=off`.
|
|
96
|
+
|
|
97
|
+
A step is state-changing when its own words — not its quoted arguments — carry one of these verbs: click, press, fill, submit, select, check, uncheck, choose, attach, upload, reload, move back / forward, go to, navigate, on the homepage, open, close, add, remove, insert, delete, save, publish, edit, update, create, enter, type, clear, reset, drag, drop, hover, tap, toggle, expand, collapse, search, send, request, flush, confirm, cancel, accept, dismiss, log in / out, sign in / out. The list covers the built-in steps and the way a project's own steps are usually worded (`When I add "Laptop" to the cart`, `When I log out`, `When I enter the 2FA code "123456"`). `Then I should see "Delete"` is no action: the verb is in the quoted text. A scroll gets no settle on purpose: it often starts a loader ("Loading more items…") that the next step checks while it shows. A project step whose words carry none of the verbs waits the same way by calling `await this.smartSettle()` itself.
|
|
93
98
|
|
|
94
99
|
## What this kills
|
|
95
100
|
|
|
@@ -25,6 +25,8 @@ Scenario: Layout assertions
|
|
|
25
25
|
Then "main" should be visible
|
|
26
26
|
```
|
|
27
27
|
|
|
28
|
+
The table may start with a row that names its columns, `| name | css selector |` (or `| name | xpath selector |` for `I define xpath selectors:`): that first row is read as the header of the table and registers nothing.
|
|
29
|
+
|
|
28
30
|
### 3. JSON file (project-wide preset)
|
|
29
31
|
|
|
30
32
|
```js
|
|
@@ -37,6 +39,8 @@ worldParameters: {
|
|
|
37
39
|
}
|
|
38
40
|
```
|
|
39
41
|
|
|
42
|
+
A file of `files` (and of `When I add selectors from "…" file`) is looked for in the `filesPath` folder first (default `./tests/selectors/`), then as written from the project root: `'auth.json'` and `'tests/selectors/auth.json'` name the same file. A file found in neither place fails the scenario with the paths tried.
|
|
43
|
+
|
|
40
44
|
Every scenario starts with a fresh registry. Later entries with the same name override earlier ones (last wins).
|
|
41
45
|
|
|
42
46
|
## Built-in CMS / framework presets
|
|
@@ -93,3 +97,23 @@ To keep tests portable, every preset exposes the same core names. See `tests/sel
|
|
|
93
97
|
- Modifier first (`active tab`, `danger button`, `first list item`).
|
|
94
98
|
- Synonyms collapse to canonical names (`alert success` / `notification success` / `callout success` → `notice success`).
|
|
95
99
|
- `data table` for HTML tables; `data list` for `ul`/`ol` collections.
|
|
100
|
+
|
|
101
|
+
## Bare names and your own steps
|
|
102
|
+
|
|
103
|
+
The steps that take registered names without quotes (`When I click primary button`, `Then I see visible header, footer`, `Then I don't see promo banner`, `Then I see header above footer`, `Then I see search input has focus`) match any unquoted words after their verb, so that every registered name works, `sign in button` and `log in link` included. A custom step of your project that starts the same way with unquoted words matches them too: `When I click next button in tour` is ambiguous between your step and the bare-name click, and cucumber-js stops at it.
|
|
104
|
+
|
|
105
|
+
Write such a step with a quoted argument and words of its own around it (`When I click "Next" in the tour`), or start it with other words (`When I move to the next step of the tour`). Or register the element and use the built-in step: `When I add "tour next button" selector for ".tour .next" css selector`, then `When I click tour next button`. After adding a step, `npx cucumber-js --dry-run` lists every step that is ambiguous.
|
|
106
|
+
|
|
107
|
+
## Which steps read a registered name
|
|
108
|
+
|
|
109
|
+
The registry is read by the bare-name steps above and by the modal steps. Every step that takes its argument in quotes reads it as it is written: a text (`I click "Subscribe"`, `I press "Save"`), a label, a placeholder or a name (`I fill in "Email" with "…"`), or a CSS selector (`"#dashboard" should be visible`, `the element "#hero" should be displayed`). A registered name given to one of those steps matches nothing, and its failure says so and names the step to use:
|
|
110
|
+
|
|
111
|
+
```
|
|
112
|
+
✖ Then "premium-banner" should be visible
|
|
113
|
+
Expected "premium-banner" to be visible.
|
|
114
|
+
Found: last seen: false
|
|
115
|
+
Page: http://localhost:8080/registered-names.html
|
|
116
|
+
Hint: "premium-banner" is a name of the selector registry (.promo-banner), but this step does not read the registry. Use: I see visible premium-banner, or: ".promo-banner" should be visible
|
|
117
|
+
```
|
|
118
|
+
|
|
119
|
+
The hint is added to the failures of the click and press steps, the fill steps, the web-first steps (`"…" should be visible`, `… should have text`, …) and the element steps (`I click on the element "…"`, `I scroll to the element "…"`, `the element "…" should be at the top of the viewport`, …).
|