0 Fork 0

mirror of https://github.com/webstudio-is/webstudio.git synced 2026-10-28 23:57:17 +09:00

Files

T

Oleg Isonen dd99d27e5e fix: Stop returning unqueryable unpublish jobs (#6588 )

Unpublish finishes in the request, but it returned a job ID that
`get-publish-job` could never find. It now returns the domains it
unpublished. A missing publish job also stays a normal not-found error
instead of suggesting a CLI update.

Closes #6578

## Changes
- Keep filename-based E2E shard discovery. Rebalance the files across
ten shards using measured CI times; each shard keeps its own test
environment and the same 12-minute safety timeout.
- Remove the fabricated unpublish job ID. `get-publish-job` remains for
publish jobs.
- Remove publish-only flags from the unpublish CLI and clarify CLI/MCP
guidance.
- Keep legacy optional server inputs for compatibility. The API input
contract does not change; clients that read the old `jobId` result need
to use `unpublishedDomains` instead.
- No database migration. Existing domain-removal behavior is unchanged,
including retention of other published domains.

## Checklist
- [x] Reproduced both reported failures with tests before the fix.
- [x] Reviewed the change for compatibility, redundant code, and
retained-domain behavior.
- [x] Updated the authoritative MCP docs and regenerated the CLI/MCP
docs.
- [x] Ran focused Builder, CLI, domain, and build tests; workspace
typecheck, lint, package-boundary, generated API, and generated-doc
checks.
- [x] Restore filename-based shard discovery and rebalance the files
across ten shards.
- [x] Verify every Chromium workflow test runs in exactly one shard and
update the E2E contributor docs. All 141 tests are assigned once across
the ten shards.
- [x] Verify parallel-safe long E2E files and the two-submission legacy
form test in CI.
- [x] Let independent slow E2E files run their tests concurrently; all
ten shard jobs passed.
- [ ] Confirm all E2E jobs finish in under five minutes. Eight did;
shard 1 took 5:10 and shard 4 took 6:32. Move the long Content-mode page
and shared-project bridge files to separate shards, then remeasure.
- [x] Required CI checks pass.

No live unpublish was run because it would change a published site.
Agent evaluations were not run; they require separate approval.

2026-10-23 18:59:44 +02:00

4.3 KiB

Raw Permalink Blame History

Builder E2E tests

Builder E2E workflows run with Playwright Test. The Playwright setup project waits for PostgREST, resets the disposable database, and warms the login route before Chromium runs the selected tests.

Run the suite

Use the full command for CI-like validation:

pnpm e2e:builder

The first run after migrations change applies the regular migration pipeline and writes an ignored schema cache under .cache/builder-e2e. Refresh it explicitly with:

pnpm e2e:builder:refresh-schema-cache

The regular E2E suite uses lightweight PostgreSQL containers. Verify that the complete migration chain also applies to the pinned Supabase PostgreSQL image with:

pnpm e2e:builder:check-supabase-migrations

Pass normal Playwright Test arguments after --. For example:

pnpm e2e:builder -- --grep "Builder can copy, duplicate, and delete a page"
pnpm e2e:builder -- 'apps/builder/e2e/tests/pages-actions.[shard-3].e2e.ts'
pnpm e2e:builder -- --debug --grep "Builder can copy"

A lightweight Playwright discovery pass validates the selection before Docker, database migrations, browser installation, or the Builder build starts. A filter that matches no tests therefore fails without starting the test environment.

Fast local reruns

Keep the E2E backend and Builder development server running to avoid rebuilding Builder and restarting Docker for each edit.

Terminal 1:

pnpm e2e:builder:dev:backend

Terminal 2:

pnpm dev

Terminal 3:

pnpm e2e:builder:dev -- --grep "Builder can copy, duplicate, and delete a page"

pnpm e2e:builder:dev defaults to E2E_BUILDER_URL=https://127.0.0.1:3000. Override it when the development server uses another port. Chromium maps wstd.dev and its project subdomains to loopback for local Builder URLs, so host-file changes are unnecessary.

The development backend uses E2E_DB_BOOTSTRAP=if-empty and keeps Docker containers running. The Playwright setup project still resets test data before each invocation.

Test organization

Put workflows in e2e/tests/*.e2e.ts and import test from e2e/test.ts.
Use Playwright's page fixture for the workflow under test. Playwright gives each test an isolated context and closes it automatically.
Use Playwright's context fixture when setup should share authentication with the workflow page.
Use withBrowserContext(browser, callback) for isolated setup or setup that runs in beforeAll. If cookies must survive from beforeAll into the tests, create an explicit context with newBrowserContext(browser) and close it in afterAll.
Use newIsolatedPage(browser) for identities that must not share cookies.
Keep tests in a file independent even though Playwright runs them in declaration order by default.

CI discovers isolated Playwright shards from filename tags. Each shard uses four workers, and files are assigned by measured CI duration to keep jobs near three to four minutes:

pages-actions.[shard-3].e2e.ts

Every E2E filename must contain at least one unique shard tag. CI derives its matrix from those tags, so adding or removing a shard does not require editing the workflow. Files with multiple tags are partitioned across those shards. All files selected by a shard must use the same set of tags. Rebalance by changing filename tags when measured job durations drift. Suites may opt into parallel mode only when every worker creates uniquely named setup data and the tests do not depend on each other's mutations.

Backend organization

Development, E2E, schema-cache, and Supabase compatibility commands share the backend definition in apps/builder/backend. compose.yaml defines the common services, and compose.<mode>.yaml defines only the differences for a named mode. Backend commands resolve these files by convention; callers do not pass Compose filenames.

The schema-cache fingerprint discovers migration sources and non-test backend files automatically. Adding a database image, initialization script, Compose override, or migration therefore invalidates the cache without updating a filename registry.

Failures

Failed shard jobs upload the Playwright HTML report, traces, screenshots, videos, and backend service logs.

When you need a completely clean backend locally:

E2E_SKIP_CLEANUP=false pnpm e2e:builder

4.3 KiB Raw Permalink Blame History