Chore: Simplify dataflow in linter.verify

[not-an-aardvark]

What is the purpose of this pull request? (put an "X" next to item)

[x] Other, please explain:

What changes did you make? (Give an overview)

This refactors linter.verify to simplify the dataflow throughout the function, and separate out some parts when possible. Specifically:

• Rules are now run as part of a separate function
• Environments are now resolved once, as opposed to several times in different places for parserOptions and globals.
• The defaults for linter.verify are now processed into a consistent object in a separate function.

This should make it easier to implement #9998.

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

Nothing in particular

[not-an-aardvark]

I removed this test because we were doing config validation in two separate places: ConfigValidator which verifies all parts of a config including rule schemas, and Linter where we were only checking if a rule's configured value was null and skipping all other validation. The second validation is generally redundant with the first validation.

[platinumazure]

LGTM, thanks!

Just out of curiosity, what was the effect of the incorrect configuration in conf/environments.js?

[not-an-aardvark]

I think it meant the builtin environment would not work when used in a config file, because it was specified differently from other environments. However, we were always adding the globals from the builtin environment anyway, regardless of configuration.