Repository navigation
Conversation
nock 14 and the global MSW server share one @mswjs/interceptors instance, so every request reaches both listeners. nock applies on import and answers first; MSW then warns about a request nock already served. A request nock has no interceptor for at all fell through MSW's warn-and-passthrough to real DNS. Add mockOriginsWithNock() and call it from every nock suite. It disables nock net connect (loopback excepted), so an unmocked request fails inside nock in milliseconds, and registers an MSW passthrough handler per origin, so MSW defers to nock whichever listener runs first and stops printing warnings. Correct the llms.txt comments: unmocked probes fail as nock errors and are retried, they do not wait on DNS.
There was a problem hiding this comment.
Copilot review overview
🟡 Changes recommended
The global Nock connection restriction is not restored during suite teardown and can leak into later test files.
Review effort: Balanced
Findings: 1
Open (1)
What changed in this PR
Adds shared network isolation for Nock-based E2E suites, preventing accidental external requests and suppressing misleading MSW warnings.
Changes:
- Adds
mockOriginsWithNock()for Nock isolation and MSW passthrough. - Applies the helper across five E2E suites.
- Documents the testing convention and corrects retry comments.
| File | Description |
|---|---|
test/nock-helpers.ts |
Adds shared Nock/MSW setup. |
test/unprocessable-content-e2e.test.ts |
Enables isolated Nock networking. |
test/scrape-progress-e2e.test.ts |
Enables isolated Nock networking. |
test/refresh-pipeline-e2e.test.ts |
Covers standard and allowlist origins. |
test/markdown-identity-e2e.test.ts |
Enables isolated Nock networking. |
test/empty-page-refresh-e2e.test.ts |
Enables isolated Nock networking. |
AGENTS.md |
Documents the Nock suite convention. |
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
Comment on lines
+36
to
+37
| nock.disableNetConnect(); | ||
| nock.enableNetConnect(LOOPBACK_HOST); |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
The nock-based end-to-end suites could send a request to the real network, and they printed misleading MSW warnings for every request. This change makes nock the only thing that answers requests for each suite's test hosts.
nock 14 and the global MSW server (
test/setup-e2e.ts) both use@mswjs/interceptors, and they share one interceptor instance, so every request reaches both of them. nock is applied when the test file imports it, so its listener always runs first and answers. MSW's listener runs after that and warns "intercepted a request without a matching request handler", even though nock already answered. When nock had no mock registered for a host at all, it let the request go, MSW passed it through, and it reached real DNS.Changes
mockOriginsWithNock()intest/nock-helpers.ts. It:beforeEach:unprocessable-content,scrape-progress,empty-page-refresh,markdown-identityandrefresh-pipeline(including its allowlist hosts).AGENTS.mdhas a short note for people writing new nock suites.About the CI timeout in run 37244567007
This does not fix the 60-second timeout in "does not let an asset-heavy site trip the failure threshold", and I don't want this pull request to suggest it does. The MSW warnings in that log look like MSW took the requests, but they show up in passing runs too: 28 per run of this suite with
--reporter=verbose. A request that really went to the network would have failed at once withENOTFOUND, which the fetcher does not retry, and the crawl would have moved on to/two. Instead the log stops right after nock answered/one. The next step for that page is the Playwright render, because the scrape mode defaults to auto. Run 35456512541 (refresh-pipeline-e2e) stalled in the same way, right after/. I could not reproduce the stall locally, so the cause is still open.Testing
npx vitest run test/unprocessable-content-e2e.test.ts: 20 runs in a row, all passedENOTFOUND); with it, the request failed in nock in 2 to 3 ms, mocked requests still worked, and loopback was still reachablenpm test: 144 files, 2325 tests passrm -f .tsbuildinfo && npm run typecheck: clean