fix(docs): resolve metadata paths with zero-width spaces

[danyalahmed1995]

Summary

Fixes #12005.

This PR resolves a docs metadata path mismatch for docs paths containing U+200B ZERO WIDTH SPACE, using the exact dogfood repro from #12052.

Changes

• Adds the exact Repro: test docs with folders containing zero-width files #12052 dogfood repro pages for zero-width spaces at the start, middle, and end of a path segment.
• Resolves docs metadata import paths back to the loaded docs source before hashing.
• Adds a small shared metadata source resolver in @docusaurus/utils.
• Applies the resolver only in the docs plugin for this PR.
• Removes the previous helper-only test.

Validation

Ran:

yarn workspace @docusaurus/utils build
yarn workspace @docusaurus/plugin-content-docs build
yarn test packages/docusaurus-plugin-content-docs/src/__tests__/index.test.ts
yarn test packages/docusaurus-plugin-content-docs
yarn build:website:fast

Manual fail/pass check:

• exact Repro: test docs with folders containing zero-width files #12052 repro + metadata-source resolution disabled: yarn build:website:fast fails with missing metadata JSON imports
• exact Repro: test docs with folders containing zero-width files #12052 repro + fix restored: yarn build:website:fast passes

No rspackBundler, rspackPersistentCache, website config, package, or lockfile changes are included.

[netlify]

✅ [V2]

Built without sensitive environment variables

Name	Link
🔨 Latest commit	044b0a5
🔍 Latest deploy log	https://app.netlify.com/projects/docusaurus-2/deploys/6a17a0619db96700089b7ef3
😎 Deploy Preview	https://deploy-preview-12044--docusaurus-2.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[slorber]

Honestly, this looks like AI slop

Maybe it's a legit fix, but you are not testing the problem end-to-end, and only adding tests for the code you introduced, which doesn't prove anything in practice.

It remains to be proved that the problem is actually solved, and that your PR is not introducing useless code, despite it being "tested."

---

Before even thinking about implementing a solution, I want to see an end-to-end test case that actually fails on our CI.

Please add a dogfooding test case to our own website, for example, as I did here: #12052

• If I remove your code, the CI must fail with the initial bug report error.
• If I add back your code, the CI must pass, proving that your implementation solves the problem.

Also, this issue is likely to also affect other plugins, so we'd rather implement a generic solution.

[danyalahmed1995]

Honestly yeah i got invested in proving my approach instead of looking at the original end to end failure , I will keep that in mind apologies for that 😅.

I am gonna do the testing you asked me to do and come back with a generic solution that doesn't affect other plugins.

Thank you for the feedback.

[danyalahmed1995]

@slorber I reworked the PR around the dogfood repro path from #12052.

Locally, with the dogfood vulnerability-report<U+200B><U+200B>/index.mdx page added:

• removing the implementation fix makes yarn build:website:fast fail with the missing generated metadata JSON import
• restoring the fix makes the same command pass

I removed the helper-only test coverage because that was basically redundant and was mainly for proving the approach not the actual repro.

I havent made a commit for it because i am still looking for a generic approach as you asked so it doesnt effect other plugins.

[slorber]

Please use the exact same dogfood test case that I added here: #12052

We have proof that it reproduces the problem with the other PR, and we'll merge it as part of the fix to prevent any possible regression in the future.

For the e2e test failures, it's an unrelated problem: just rebase your PR against main.

[danyalahmed1995]

I tested the exact #12052 dogfood repro in a clean sibling worktree instead of my earlier custom repro.

With the default Rspack config, yarn build:website:fast fails with the missing metadata JSON imports for the zero-width-space docs paths.

After I changed only:

rspackBundler: false,
rspackPersistentCache: false,

cleaned website/.docusaurus, website/build, and cache dirs, and the same command passed.

So yeah, this looks Rspack-specific. I’ll rebase against main and update the PR around the exact #12052 dogfood test case, not the helper-only test.

[danyalahmed1995]

Updated the PR after rebasing against main.

The diff now uses the exact #12052 dogfood repro files, removes the helper-only test, and keeps the fix scoped to the docs plugin while using a small shared metadata source resolver.

Local validation confirms:

• exact Repro: test docs with folders containing zero-width files #12052 repro + metadata-source resolution disabled: yarn build:website:fast fails with missing metadata JSON imports
• exact Repro: test docs with folders containing zero-width files #12052 repro + fix restored: yarn build:website:fast passes

No Rspack config, website config, package, or lockfile changes are included.

[slorber]

If this is a Rspack bug, then the fix must be reported with a minimal repro to the Rspack team

I don't want to implement a weird workaround for a bug in a dependency, to solve a fancy edge case (who uses zero-width spaces in filenames 🤷‍♂️ )

[danyalahmed1995]

Well honestly this issue seemed weird to me from the start , i picked it because had a feeling that there is a lot more to this than something that's being reported but i am still happy with the findings we got it at least gives a ground to work on future if need . Not to mention it also got super complicated I should have talked to you to understand if the PR is needed. I will keep that in mind for future 😄