Cmd no globals

[mstoykov]

edit: go commit by commit otherwise it's very hard to follow what happens IMO
edit2: I have optimized for small change so the code review can be done faster and also to not break anything. The final code is still pretty terrible for many ... many reasons, but fixing those is IMO outside the scope of this PR.

edit3: this was split so that the smaller changes are now in #2357 and #2358. And this PR is based on #2358 as it kind of depends on it to run the tests in parallel

[olegbespalov]

A PR looks good 👍

However, I have a small readability improvement point. What if we use a more explicit name instead of the fl?

At some point, we (and other people) are going to read it again and ideally, we should limit the number of the shortcuts like that.

And sometimes it looks not consistent, example:

outputs, err := createOutputs(conf.Out, src, conf, runtimeOptions, executionPlan, osEnvironment, logger, fl)

[mstoykov]

However, I have a small readability improvement point. What if we use a more explicit name instead of the fl?

as long as you come up with something. I guess part of the review process will be coming up with names about commandFlags and fl as both are ... bad.

Also I might break this in 3 PRs :

1. the first commit here
2. the last commit here without the paralization
3. everything else - which IMO has the most to be discussed about

WDYT?

[mstoykov]

All of those os.Getenv should go, but as currently this are the only places where they are and there are zero tests covering this I decided against introducing even more changes.

[codebien]

an amount of code that I hope at some point will be rid by croconf, just spot one tiny typo

[codebien]

Suggested change

	// TODO better name - there are other commadn flags these are just ... non lib.Options ones :shrug:
	// TODO better name - there are other command flags these are just ... non lib.Options ones :shrug:

[mstoykov]

I've update the name and have split this in 3 PRs so this is only about the flags

[olegbespalov]

Sorry for the delay, PR looks good to me! 💪

[mstoykov]

Sorry for the delay

Don't worry, I actually will be waiting for @na-- to give a 👍 as well, as the person who probably is most familiar with all the things that can go wrong in this code

[na--]

LGTM, besides the minor nitpick for the comment! 🎉 This is definitely not the final state these should be in, but a great step in the right direction! ❤️