deepStrictEqual and notDeepStrictEqual do not compare Error causes nor AggregateError errors arrays

[Zamralik]

Version

v20.9.0

Platform

Linux DESKTOP-8H17SD6 5.15.133.1-microsoft-standard-WSL2 #1 SMP Thu Oct 5 21:02:42 UTC 2023 x86_64 GNU/Linux

Subsystem

No response

What steps will reproduce the bug?

• Compare 2 Error one with a cause and one without
• Compare 2 Error one with different causes
• Compare 2 AggregateError with different array of errors

These 4 calls should not throw.
Inversely, replacing notDeepStrictEqual by deepStrictEqual should throw for all.

import { notDeepStrictEqual } from "node:assert";

notDeepStrictEqual(
    new Error('Test'),
    new Error('Test', { cause: new Error('Expected Cause') }),
);

notDeepStrictEqual(
    new Error('Test', { cause: new Error('Unexpected Cause') }),
    new Error('Test'),
);

notDeepStrictEqual(
    new Error('Test', { cause: new Error('Unexpected Cause') }),
    new Error('Test', { cause: new Error('Expected Cause') }),
);

notDeepStrictEqual(
    new AggregateError([], 'Aggregate Test'),
    new AggregateError([new Error('Expected Child')], 'Aggregate Test'),
);

notDeepStrictEqual(
    new AggregateError([new Error('Unexpected Child')], 'Aggregate Test'),
    new AggregateError([], 'Aggregate Test'),
);

notDeepStrictEqual(
    new AggregateError([new Error('Unexpected Child')], 'Aggregate Test'),
    new AggregateError([new Error('Expected Child')], 'Aggregate Test'),
);

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

Every time

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

notDeepStrictEqual should throw when the Error's cause / AggregateError's errors match perfectly recursively.
deepStrictEqual should throw when the Error's cause / AggregateError's errors do not match perfectly recursively.

What do you see instead?

notDeepStrictEqual throw when the Error's cause / AggregateError's errors match perfectly recursively.
deepStrictEqual do not throw when the Error's cause / AggregateError's errors do not match perfectly recursively.

Additional information

No response

[kylo5aby]

As mentioned in deepstrictequalactual, for Error type, it compares names, messages and enumerable "own" properties.

cause, message, and name, are both non-enumerable properties. However, as mentioned in ECMA2024, cause triggers some special operations during the creation of an Error instance, so I believe cause serves as another distinguishing attribute among different Error, much like message and name. I'm unsure if others agree with this perspective, and I'm willing to submit a PR if feasible.

[Zamralik]

I use AggregateError and Error's cause to create a tree to provide an explicit trail of what was wrong and where.

Being able to recursively compare the Error's cause/Aggregate's errors in the tests is more thorough that no info on the reason is missing.

[BridgeAR]

I agree with @zhenweijin that we should just add a special handling for it. It should be pretty straight forward to add that check.

[abidjappie]

I'm interested in looking at this issue if it become available 🙇 (I see it's still in Triage)

[Elham-EN]

Hi, can you assign me to this issue?

[BridgeAR]

There's an open PR, so it's almost done.