[v8.x backport] util: expand test coverage for util.deprecate

[akaila]

Backporting: #16305

Checklist

• make -j4 test (UNIX), or vcbuild test (Windows) passes

Affected core subsystem(s)

util

[akaila]

Merge in other dependent changes as well.

[trivikr]

@akaila Can you rebase instead of merge?
Details: https://github.com/nodejs/node/blob/63243bcb330408d511b3945c53719425d8b7abb8/deps/uv/CONTRIBUTING.md#rebase

[MylesBorins]

@akaila where did you get 8ea3698 from

[akaila]

@trivikr I did a cherry pick as required in backporting. This is not a merge or rebase.
@MylesBorins I had to make a new commit (8ea3698) since global expect error fails the test if SSL is present. I don't know how or if this is passing in master.

[akaila]

FYI this is the test failure without the fix:

python tools/test.py -J --mode=release parallel/test-internal-util-assertCrypto
=== release test-internal-util-assertCrypto ===
Path: parallel/test-internal-util-assertCrypto
Mismatched function calls. Expected exactly 1, actual 0.
at Object.exports.mustCall (/Users/kaila/Developer/Node/test/common/index.js:499:10)
at Object.expectsError (/Users/kaila/Developer/Node/test/common/index.js:724:27)
at Object. (/Users/kaila/Developer/Node/test/parallel/test-internal-util-assertCrypto.js:7:30)
at Module._compile (module.js:612:30)
at Object.Module._extensions..js (module.js:623:10)
at Module.load (module.js:531:32)
at tryModuleLoad (module.js:494:12)
at Function.Module._load (module.js:486:3)
at Function.Module.runMain (module.js:653:10)

[apapirovski]

You shouldn't include anything that is not directly related to the specific PR you're backporting, even if tests are failing. That can be solved in a separate PR if necessary. That said, error changes are semver-major so that fix doesn't seem applicable to v8.x anyway.

[akaila]

Ok I took out that commit. The only commit now is my test and its dependent code.

[MylesBorins]

@apapirovski the error change in 8ea3698 is in a test, not in core... so it wouldn't be semver major

I too am very confused as to how that is passing on master

edit: @apapirovski my bad, I'm seeing that the change to error codes is in the backport.

[apapirovski]

@MylesBorins The issue I'm seeing is with a07981d811e2c22c57dbaf614a1caa94a34f30d4 being included here

[MylesBorins]

@akaila we cannot land a07981d in v8.x as it is a semver major change. As it would appear that the backport requires the error change I'm going to go ahead and close this

My apologies for the miscommunication, thank you so much for taking the time to open this

[apapirovski]

@MylesBorins I think this all got very confused. 😭 This should still be open, the issue is that it includes a07981d (incorrect) alongside 82134aa (correct)

Maybe I'm missing something though...

[apapirovski]

The real solution here should be to backport that test but change the error being tested for, since the error is different.

`code` argument must be a string

Without using the internal errors.

[akaila]

I get it now. I will revert and modify the test and update the PR. Thanks

[MylesBorins]

hey all... common.expectsError is only for testing internal/error objects... as such I don't think this makes sense to continue with. It will be in 9.x, and the expanded coverage doesn't make sense for 8.

I will avoid closing again but I advise that we close this

[targos]

common.expectsError should work with plain Errors too.

[apapirovski]

@MylesBorins Is there downside to having an extra test for util.deprecate in v8.x? Even if common.expectsError doesn't work (I think it should?), we can still use assert.throws...

[MylesBorins]

@targos gtk
@apapirovski no downside

Sorry for being a party pooper, please safely ignore me

[akaila]

Backported the test and corrected the error message and removed the error code. Verified it passes.

[gibfahn]

@akaila thanks a lot for sticking with this!

Landed as 55ba1d4