Hey @mheiber! We really appreciate you taking the time to report an issue. The collaborators
on this project attempt to help as many people as possible, but we're a limited number of volunteers,
so it's possible this won't be addressed swiftly.

If you need any help, or just have general Babel or JavaScript questions, we have a vibrant Slack
community that typically always has someone willing to help. You can sign-up here
for an invite.

It seems that V8 will print #foo as the function name instead of foo.

class A {
    #foo = function () {};
    constructor () {
        console.log(this.#foo.name, this.#foo.name.length);
    }
}

new A();

prints #foo 4 instead of foo 3 according to the spec.

I have submitted a bug report to V8: https://bugs.chromium.org/p/v8/issues/detail?id=9481

Hi @JLHwung whoops! I think I made a mistake earlier. Sorry about that.

It looks to me now that the function.name should be #foo:

https://tc39.es/proposal-class-fields/#sec-runtime-semantics-evaluate-name

lexical semantics:

• privateIdentifier

runtime semantics
names are created with newPrivateName(privateIdentifier)

in the definition of newPrivateName(description):
- [[description]] is set to description

in the lexical grammar:

• privateIdentifier :: # identifierName

Does this seem to be a real issue, then? I could make a PR in a few weeks after I get approval from my company to contribute.

@mheiber It is a bug and PR is definitely appreciated.

Note that Babel has similar issues on keeping track of .name property, (i.e. #7425 and #9562). Most of this issue can be solved by guess a name from AST and use it as the BindingIdentifier name of the function declaration.

However, as # is not a valid character for identifier name. This one will be more difficult to tackle with than those I mention above. We may have to append Object.defineProperty calls to the function declaration.

I would prefer not to use defineProperty and to give the name without the hash. Function names are one of the parts of the spec where we aren't really compliant, since what matters most is that they are usable for debugging.

Function names are one of the parts of the spec where we aren't really compliant, since what matters most is that they are usable for debugging.

I agree, at least it will be better than our current situation, though I am afraid it might confuse developers if we return the name without hash while native implementations return the name with hash, which is spec compliant. Maybe we can leave a note comment on that.