[WIP] Parallel lint test

[timocov]

PR checklist

• Addresses an existing issue: Run linting logic in spawned threads #2328
• Enhancement
  • Includes tests
• Documentation update

Overview of change:

It's the first version of parallel linting. See #2328 (comment).

The main changes are:

1. Slight refactoring around runner (moved some interfaces to separate files, added/created ability to filter rules typed/non-typed, etc) and running linting (moved some code to src/doLinting.ts module, etc).
2. Added new CLI option --parallel to run linter in parallel.

Is there anything you'd like reviewers to focus on?

1. The whole PR 🙂
2. Grammar (English is not my language).
3. Files/classes/types names - I'm not familiar with tslint code base a lot and might misunderstood something.
4. I've place TODOs in some places with comments
5. How we can test it? 🤔

CHANGELOG.md entry:

[timocov]

In my local tests with running parallel linting for tslint code base I got decreasing lint time from 10 to 7 seconds on my PC with 2 workers. If I run with 8 workers (I have i7 with 4 cores/8 threads), the parallel run is a little bit slower (~1 second) - I guess that is because spawning the process isn't cheap and we need to provide users specify how much workers they want to have.

[timocov]

Hey @JoshuaKGoldberg, I don't know whether we're going to review/merge this, but I just merged master into the branch and resolved conflicts 🙂

[JoshuaKGoldberg]

Thanks @timocov! It's still unclear what's the right course to take. /cc @adidahiya

[JoshuaKGoldberg]

Ping @adidahiya - I think this would be a reasonable change to include in v6 (#4811) if documented as an 'experimental' option. If there are severe bugs with it, well, TSLint's going away anyway. Thoughts?

[JoshuaKGoldberg]

@timocov I'm going to go ahead and close this PR. We have, on average, about half a person's worth of maintenance on the TSLint side (no full time maintainers), and we're facing delays just getting the breaking change issues merged in. It doesn't seem feasible to add this feature given how late we are in the ESLint migration process: #4534. 😢. If something were to go wrong we can't guarantee there'd be staffing to fix it and release that fix.

Still, to reiterate: the code change itself looks great and I think you did a fantastic job on it. You can always make a forked version of TSLint to use locally - extra features being added to TSLint isn't too big of a concern anymore.

Thanks, again for taking the time to work with us on this ❤️. Sorry it didn't work out ...!