feat: init nodejs-helpers package, add bypass-import-map, bypass-impo…

[narzac]

…rt-map tasks, and run part of lion prepublish

What I did

1. init nodejs-helpers package
2. add bypass-import-map
3. add bypass-export-map
4. run part of lion prepublish

[changeset-bot]

🦋 Changeset detected

Latest commit: 122a481

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 2 packages

Name	Type
@lion/nodejs-helpers	Patch
@lion/ui	Minor

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[CLAassistant]

All committers have signed the CLA.

[tlouisse]

Did you consider using mock-fs (for readability and perf)?

Then we can do:

mockFs({
  '/simple-export-map/components/icon/xyz.js': "export x = () => 'bla';"
  // ... etc
})

[narzac]

Yes I did consider mocking filesystem, however it is another layer of abstraction on top of the reality, which will make it less readable imo, and more of a complex solution. Of course, as you suggested we could still consider it, say for performance, and that would certainly be the case, if we were mocking timers and so on, yet performance wise the whole tests that access the filesystem are taking ~0.247 seconds to be precise.

To get the ~0.247, we can

First run the tests without the tests which uses the filesystem

rm -rf packages-node/nodejs-helpers/test-node/tasks/ && time npm run test -w packages-node/nodejs-helpers/

real 0m1.323s

And then, run it with them:
g checkout packages-node/nodejs-helpers/test-node/tasks/ && time npm run test -w packages-node/nodejs-helpers/

real 0m1.570s

Then the difference is ~0.247,, which is insignificant.

[tlouisse]

For a future improvement: maybe we can leverage globby's gitignore option?

[tlouisse]

At some point, we might also support css import assertions. At that point, we can probably extract the generic export map resolver from providence and use it here as well.