Add option and CLI flag to hide command output

[firoxer]

Hi,

I noticed #138 and thought it might be doable. Here's a proof-of-concept that does the job as proposed by @gustavohenke in the aforementioned issue.

This is not a complete PR because some edge cases might be missing and the documentation is incomplete for this part. I'll get to them if I get an OK on the POC.

Issue #25 is affected by this PR as well, as the proposed --hide flag allows for implementing --silent/-s more easily.

Thanks for creating concurrently!

[coveralls]

Pull Request Test Coverage Report for Build 325

• 3 of 3 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.02%) to 99.198%

---

Totals
Change from base Build 323:	0.02%
Covered Lines:	242
Relevant Lines:	242

---

💛 - Coveralls

[coveralls]

Pull Request Test Coverage Report for Build 325

• 3 of 3 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.02%) to 99.198%

---

Totals
Change from base Build 323:	0.02%
Covered Lines:	242
Relevant Lines:	242

---

💛 - Coveralls

[coveralls]

Coverage remained the same at 100.0% when pulling a41913e on firoxer:add-option-to-hide-output into fc2817a on kimmobrunfeldt:master.

[gustavohenke]

This PR is awesome! Thanks for taking the stab!

I'm curious about this:

This is not a complete PR because some edge cases might be missing and the documentation is incomplete for this part

What else do you think should be covered by your PR?

I only miss some functional tests, that make sure the CLI flag will work,
but I'm happy to know your input!

[firoxer]

What else do you think should be covered by your PR?

I was just pointing out that I'm not confident that all use cases have been covered. :) If you feel it's complete then it's all good.

I added some functional tests and they seem to work fine. While doing that, I noticed that because --hide uses the yargs array type, it makes it necessary to use -- after the final hideable index/name when --hide is the last option because otherwise yargs won't know where stop parsing indices/names and where to start parsing commands. In other words, if the first and the third process of bin/concurrently.js 'echo 1' 'echo 2' 'echo 3' should be hidden, the command should look like bin/concurrently.js --hide 0 2 -- 'echo 1' 'echo 2' 'echo 3'. I expanded the usage notes to include this.

However, I was left wondering if this kind of behaviour is fine, because it kind of conflits with --names which works with comma-separated values instead of spaces (e.g. --names foo,bar, no following -- needed). The problem is in that by telling yargs to use array, we get automatic casting to numbers (i.e. --hide 1 foo 2 bar results in an array of [1, 'foo', 2, 'bar']) which works really well with the rest of the code. By making --hide work like --names we lose this nicety, but we gain usage consistence. Which way do you prefer?

[AntonNiklasson]

Awesome job, I would love to see this merged. However, I would expect the name of the flag to be "quiet" and not "hide". Thoughts?

[firoxer]

Sure. I just went with "hide" since it was suggested in #138. I guess CLI tools usually use -q or --quiet with zero arguments so --quiet [number] would deviate from that, but on the other hand I don't think it's that big of a deal. If you want me to change it, I'll do it.

[StefanFrederiksen]

This use-case is very ideal for me, so is there anything that still needs to be done for this PR?

[firoxer]

From my part I only need to know if we want to use --hide or --quiet and I'll change the code if necessary.

[StefanFrederiksen]

I think --silent is more verbose as to the intent of the flag. Otherwise --quiet since a hide flag that makes the child process not output anything doesn't seem intuitive (at least to me).

[munrocket]

Very useful PR why you not merge it @kimmobrunfeldt @gustavohenke ?

[bohdanbirdie]

@kimmobrunfeldt any chance merging this? It would be really great to have it!
Thanks!

[cben]

@firoxer can you resolve merge conflicts?

[firoxer]

@cben Sure, done.

[munrocket]

@kimmobrunfeldt @gustavohenke

[shilomagen]

This is an amazing feature we're waiting for :)
Do you plan to merge it soon? Thank you so much!

[coveralls]

Coverage increased (+0.03%) to 98.618% when pulling 863e384 on firoxer:add-option-to-hide-output into 08eda4f on open-cli-tools:master.

[gustavohenke]

Wow. Nearly 3 years later... I pushed some small changes, most notably not using the yargs array type.

LGTM now, let's ship it.

_{I feel bad that I needed the harsh COVID-19 restrictions in Sydney + a rainy long weekend for me to do this}

[gustavohenke]

Published in v6.4.0!