[nonextensible-applies-to-private] Adding private field to non-extensible object throws

[erights]

Alternative to #4563

To be the test262 tests for the https://github.com/tc39/proposal-nonextensible-applies-to-private

See tc39/proposal-nonextensible-applies-to-private#5

Initial PR content by @o-

[ptomato]

Looks correct so far.

For completeness, would we need to test stamping a private accessor and private method?

(I would guess that the case of stamping a private static method, or private static field cannot occur, since those go on the constructor, which cannot be frozen prior to the class declaration.)

Finally, this needs a feature flag for the proposal to be added in features.txt and in the frontmatter of the two modified tests.

[erights]

Looks correct so far.

For completeness, would we need to test stamping a private accessor and private method?

For inheritance, done.

For a base class, AFAICT, impossible. There are no observable states where the class, instance, or prototype are observable before methods are accessors are added.

(I would guess that the case of stamping a private static method, or private static field cannot occur, since those go on the constructor, which cannot be frozen prior to the class declaration.)

For a private static method, AFAICT, it is indeed impossible.

For a private static field, it can be done in the initializer expression. Done.

Finally, this needs a feature flag for the proposal to be added in features.txt and in the frontmatter of the two modified tests.

Done.

(with much help from @gibson042 )

PTAL

[ptomato]

As far as I can tell complete now, thanks!

[ptomato]

I left this open for a few days in case any other maintainers had any feedback, but now that it's been presented in the TC39 meeting I think it's time to merge it. I'll do that now.