Unexpected `#each` block behavior with destructuring

[x0k]

Describe the bug

Using the construct {#each gen() as [item]} with a generator that mutates the yielded object produces a result that differs from the behavior of a for…of loop with destructuring.

If this behavior is expected, it should be documented.

Reproduction

REPL

Logs

System Info

System:
    OS: Linux 6.12 cpe:/o:nixos:nixos:25.05 25.05 (Warbler)
    CPU: (12) x64 Intel(R) Core(TM) i5-10400F CPU @ 2.90GHz
    Memory: 4.58 GB / 15.56 GB
    Container: Yes
    Shell: 5.2.37 - /run/current-system/sw/bin/bash
  Binaries:
    Node: 24.7.0 - /nix/store/ffx0ixqnd6arwnjqmjv844krhay4iycj-nodejs-24.7.0/bin/node
    npm: 11.5.1 - /nix/store/ffx0ixqnd6arwnjqmjv844krhay4iycj-nodejs-24.7.0/bin/npm
    pnpm: 10.15.1 - /nix/store/dmnyvw4gf3i8b65s3k7q4ws3257m6zrs-pnpm-10.15.1/bin/pnpm
  Browsers:
    Brave Browser: 142.1.84.141
  npmPackages:
    svelte: ^5.30.0 => 5.43.14

Severity

annoyance

[7nik]

Such a generator is bug-prone because it returns the same object on each iteration. And thus you get what you get.
For example, for (const [item] of Array.from(gen())) (basically what #each is) gives the same "undesired" result.

[x0k]

This generator is really bad, since it implies usage only with destructuring, which cannot be guaranteed by normal means.

My issue was that, when using familiar syntax (destructuring), I expected similar behavior, because it’s impossible to infer the Array.from(array) behavior from the each block syntax alone.

I don’t think this behavior should be changed, especially not in ways that would hurt performance, but it would be helpful if this nuance were mentioned in the documentation.