fix: proxify values when assigning using ||=, &&= and ??= operators

[Rich-Harris]

fixes #14268 (WIP — doesn't tackle classes yet)

Before submitting the PR, please make sure you do the following

• It's really useful if your PR references an issue where it is discussed ahead of time. In many cases, features are absent for a reason. For large changes, please create an RFC: https://github.com/sveltejs/rfcs
• Prefix your PR title with feat:, fix:, chore:, or docs:.
• This message body should clearly illustrate what problems it solves.
• Ideally, include a test that fails without this PR but passes with it.

Tests and linting

• Run the tests with pnpm test and lint the project with pnpm lint

[changeset-bot]

🦋 Changeset detected

Latest commit: 4f8c987

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 1 package

Name	Type
svelte	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[github-actions]

Playground

pnpm add https://pkg.pr.new/svelte@14273

[Rich-Harris]

preview: https://svelte-dev-git-preview-svelte-14273-svelte.vercel.app/

this is an automated message

[Rich-Harris]

@trueadm can you rescind your approval? there are failing tests, and this doesn't fix the case involving public state fields

[trueadm]

Rescinding

[Rich-Harris]

So:

1. there are some tests failing on this PR that I'm not able to look into right now (if anyone else is so inclined, feel free)

2. this doesn't fix the (this.items ??= []).push(...) case, because the result of evaluating (this.items ??= []) is just [] in the case where this.items is nullish, not the proxified array that is assigned to this.items as a result of the generated set. I'm not sure we can fix this.

   We could make it work by transforming every assignment that forms part of a larger expression...

   -(a.b ??= c).push(d);
   +(a.b ??= c, a.b).push(d);

   ...but our powers stop at the .svelte[.ts|.js] boundary, and in any case it would be problematic in cases like this, requiring a more invasive transformation:

   (a().b ??= c, a().b).push(...); // we just called a() twice!

   In cases like these I think the appropriate move is to issue a warning, alerting people to the limitation.

[trueadm]

Yeah a warning makes most sense.