fix: add missing .d.cts files while building plugin-kit

[liangmiQwQ]

Prerequisites checklist

• I have read the contributing guidelines.

AI acknowledgment

• I did not use AI to generate this PR.
• (If the above is not checked) I have reviewed the AI-generated content before submitting.

What is the purpose of this pull request?

Please read sxzz/rolldown-plugin-dts#1230. ESLint should emit .d.ts for commonjs for downstream's build.

What changes did you make? (Give an overview)

Copied core package's bundling logic

Related Issues

sxzz/rolldown-plugin-dts#1230

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: liangmiQwQ / name: Liang Mi (57a9e3a, 749d75f, b732f6a)

[eslint-github-bot]

Hi @liangmiQwQ!, thanks for the Pull Request

The pull request title isn't properly formatted. We ask that you update the pull request title to match this format, as we use it to generate changelogs and automate releases.

• The commit message tag wasn't recognized. Did you mean "docs", "fix", or "feat"?
• There should be a space following the initial tag and colon, for example 'feat: Message'.

To Fix: You can fix this problem by clicking 'Edit' next to the pull request title at the top of this page.

Read more about contributing to ESLint here

[DMartens]

I cannot reproduce this problem as dist/cjs/index.d.cts for plugin-kit is created.
You can see this in the package.json script build:cts you want to modify.
Regarding the EasyCLA problem, please retry maybe in a different browser, as we cannot proceed otherwise.

[liangmiQwQ]

I cannot reproduce this problem as dist/cjs/index.d.cts for plugin-kit is created. You can see this in the package.json script build:cts you want to modify. Regarding the EasyCLA problem, please retry maybe in a different browser, as we cannot proceed otherwise.

I want to point out the file we need is the type.d.ts, which is missing in cjs directory. I think we should just do the things ESlint has already done in core package instead of modifying or calling the global building script (as there are no other imports in type.cts).

rewrite/packages/core/package.json

Line 20 in 70b6997

"build:cts": "node -e \"fs.cpSync('dist/esm/types.d.ts', 'dist/cjs/types.d.cts')\"",

And I will try to solve EasyCLA asap. Thank you.

Update: EasyCLA was solved.

[lumirlumir]

Actually, it’s a bit difficult to pinpoint the root cause here, since your reproduction link is about the rolldown integration.

Could you please share a failing minimal reproduction link without the rolldown integration?

Also, updating the PR template to use our bug template would be helpful here, so others can more easily recognize the real problem.

[liangmiQwQ]

Could you please share a failing minimal reproduction link without the rolldown integration?

Umm... this might be a little bit hard to describe or be shown in a reproduction repo, I think I need to re-explain the question.

@eslint/plugin-kit exposes index.d.(c)ts for both esm and cjs, which imports types.(c)ts. The bundler will automatically redirects all pure .(c)ts files to .d.(c)ts as a type declaration (most of the bundlers do it, like rollup, as well as rolldown).

And for the current @eslint/plugin-kit's building process, the types.d.ts file only exists under esm directory, not cjs directory, which causes bundler can't find the target .d.cts when using CommonJS.

So, for packages who depend on and bundle @eslint/plugin-kit.

• If they are using Rollup, it will automatically generate a types.d.cts as it is missing.
• If they are using rolldown (like above), it won't generate this file, make bundling error happens.

I've talked with rolldown team, they thought it is expected and it's an ESLint's side building config issue. It makes sense, because the esm side has types.d.ts (Generated by tsc), but the cjs has no types.d.cts, which is missing exactly.

Hope it's helpful, thanks.

[DMartens]

Thank you for the explanation.
I think that this is not bundler-specific problem and agree that this should be fixed on our side.
The changes make sense to me, leaving open for @lumirlumir.

[liangmiQwQ]

Thank You!