Add export fields to manifests to prevent deep imports

[Larry1123]

Todo:

• fix anything that can't resolve (likely test).
• Settle on what to do with types condition

What's the problem this PR addresses?

This is help with 4.x as requested in #3591
This addresses the task Add export fields to our manifests to prevent deep imports

How did you fix it?

Added exports to all manifests.
Added exports for some things that were not exported before.

Checklist

• I have read the Contributing Guide.

• I have set the packages that need to be released for my changes to be effective.

• I will check that all automated PR checks pass before the PR gets reviewed.

[merceyz]

I converted the PR to a draft since that's what the title said it was.

[Larry1123]

Thanks I tried to look for the button for that but I guess I just missed it.

[Larry1123]

I am not too sure how to fix scripts/test-typescript-patch.js
Some help with that would be nice.

[Larry1123]

I think that packages/acceptance-tests/pkg-tests-specs/sources/pnp-esm.test.ts also may need someone that knows the @yarnpkg/pnp package better judge how best to handle fixing the import of it in this test.
I don't clearly see that this loaderFlags should be exported so I am not too sure how to address this.

I tried adding it only to the exports field and not the publishConfig.exports but I'm not sure if that is the right way to go about this.

[Larry1123]

The only real question I have left for this PR is was flagging all affected packages as a major change right? I figured as it blocks deep imports that it likely should be counted as such.

[merceyz]

According to https://www.typescriptlang.org/docs/handbook/release-notes/typescript-4-7.html#packagejson-exports-imports-and-self-referencing we don't need the types condition since the .d.ts file is next to the .js file.

The only real question I have left for this PR is was flagging all affected packages as a major change right? I figured as it blocks deep imports that it likely should be counted as such.

Yeah, this is a breaking change so major is correct.

[Larry1123]

I have had issues when the types condition is omitted, I will confirm firsthand that it will be fine to omit the condition before removing them. In most manifests there were types or typeings in publishConfig before would those want to be removed also since they would fall under the same redundancy?

[arcanis]

I think we can go with the removes types field, and publish a rc - if there's a problem it should hopefully be spotted there

[arcanis]

Thanks for cleaning all this, btw!