Incorrect result from isDeepStrictEqual on two different Sets

[DoubleClickPreferred]

Version

22.21.1

Platform

Microsoft Windows NT 10.0.19045.0 x64

Subsystem

util

What steps will reproduce the bug?

The problem can be seen in a quick REPL session:

Welcome to Node.js v22.21.1.
Type ".help" for more information.
> a = new Set([0, new Set([1, 2, 3]), new Set([4, 5, 6])])
Set(3) { 0, Set(3) { 1, 2, 3 }, Set(3) { 4, 5, 6 } }
> b = new Set([0, new Set([1, new Set([2, 3]), new Set([20, 30])]), new Set([4, new Set([5, 6]), new Set([50, 60])])])
Set(3) {
0,
Set(3) { 1, Set(2) { 2, 3 }, Set(2) { 20, 30 } },
Set(3) { 4, Set(2) { 5, 6 }, Set(2) { 50, 60 } }
}
> require('util').isDeepStrictEqual(a, b)
true

How often does it reproduce? Is there a required condition?

Interestingly, require('util').isDeepStrictEqual(b, a) returns false, as expected. Which indicates that isDeepStrictEqual is not symmetrical.

What is the expected behavior? Why is that the expected behavior?

The result should be false.

What do you see instead?

I'm seeing true.

Additional information

No response

[ljharb]

I wouldn't expect any of those to return true, personally - new Set([1, 2]) and new Set([1, new Set(2)]) aren't the same.

[DoubleClickPreferred]

With these two inputs, isDeepStrictEqual works as expected:

Welcome to Node.js v22.21.1.
Type ".help" for more information.
> a = new Set([1, 2])
Set(2) { 1, 2 }
> b = new Set([1, new Set([2])])
Set(2) { 1, Set(1) { 2 } }
> require('util').isDeepStrictEqual(a, b)
false
> require('util').isDeepStrictEqual(b, a)
false

It looks like the bug happens with deeper sets.

[ljharb]

I mean it's clearly a bug - but "strict deep equal" shouldn't be equating new Set(xs) with xs in the first place.

[BridgeAR]

This is a case specific to mixed primitive / object cases. I will open a fix tomorrow, I identified the issue in the code. It also applies the same to maps.