Fix: prefer-destructuring error with computed properties (fixes #9784)

[not-an-aardvark]

What is the purpose of this pull request? (put an "X" next to item)

[x] Bug fix (#9784)

What changes did you make? (Give an overview)

This fixes the prefer-destructuring bug described in #9784.

Is there anything you'd like reviewers to focus on?

Nothing in particular

[platinumazure]

Not sure I understand the issue.

One could destructure like ({ [foo] } = bar), right? Or would that assign not to foo but to whatever foo holds? Or will that do an array destructure or something weird?

I'm just trying to figure out if there's ever a use case for allowing prefer-destructuring to enforce for computed properties, that's all.

[not-an-aardvark]

I don't think ({ [foo] } = bar) is valid syntax. It would be possible to use ({ [foo]: foo } = bar) here, but I think the rule should only enforce that when the enforceForRenamedProperties option is used. I think the goal of the enforceForRenamedProperties: false behavior is to enforce destructuring when it would be possible to use object shorthand (e.g. ({ foo } = bar) rather than ({ foo: foo } = bar)), which is never possible when using computed properties.