Change CLI file parsing errors to use an error code

[evaline-ju]

Description of the Change

After the implementations of #3467 and #3666 and #4464, there are a few instances in the code that do not use error codes yet.

Alternate Designs

Open to other names for this error code

Why should this be in core?

Using codes is a best practice like in Node.js, codes help identify errors as mocha errors at a glance

Benefits

Using codes is a best practice like in Node.js, codes help identify errors as mocha errors at a glance

Possible Drawbacks

None I'm aware of

Applicable issues

Contributes to #3656, semver-minor

[coveralls]

Coverage increased (+0.008%) to 94.209% when pulling 3c81ed8 on evaline-ju:issue/codes-3656-2 into d7ed5c2 on mochajs:master.

[boneskull]

thanks Evaline! I have some comments/suggestions... please let me know what you think.

[boneskull]

There's a conflict in lib/errors.js because I added a TIMEOUT error... should be able to just use both sets of changes

[boneskull]

in case anyone else was wondering, I think "unparsable" is correct and "unparseable" is not. my spell checker dislikes both 😄

[evaline-ju]

I'm not really sure if there is/was a temporary problem with the browser tests since I saw at least another PR with the same failure. At the same time, I don't know how to just re-run the test. 🤔

[boneskull]

hmm. I'll look into that error. some saucelabs weirdness.. hope it's just temporary.

this seems to have a conflict now--could you please rebase onto master?

[evaline-ju]

hmm I'm not seeing a conflict currently, after I merged from master 2 days ago?

[evaline-ju]

After a long hiatus I have tried to formally rebase - looks like the browser test on forked PRs problem is known and still existing

[outsideris]

You can ignore the browser test. We are working on #4553 .

[juergba]

@evaline-ju thank you for this PR.