fix(14815): warn when an invalid <select multiple> value is given

[GauBen]

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

Emit a warning instead of crashing the runtime when an invalid value is provided

Closes #14815

• 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.
• If this PR changes code within packages/svelte/src, add a changeset (npx changeset).

Tests and linting

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

[changeset-bot]

🦋 Changeset detected

Latest commit: 6564b41

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

[Rich-Harris]

preview: https://svelte-dev-git-preview-svelte-14816-svelte.vercel.app/

this is an automated message

[github-actions]

Playground

pnpm add https://pkg.pr.new/svelte@14816

[paoloricciuti]

Overall is good...however we recently switched to let TS handle this kind of typing issues so maybe we can just prevent the fail without throwing the warning.

In this case it's a bit different tho because this will be a much more descriptive error... I'll let some other maintainer with more experience decide on this 😁

[paoloricciuti]

I'm pretty sure this has performance implications...I don't know why it was written this way tho

[GauBen]

If we want better (theoretical) performance, we should use neither .indexOf nor .includes (code will be O(select.options.length * value.length)) but rewrite it as

const values = new Set(value);
for (var option of select.options) {
  option.selected = values.has(get_option_value(option));
}

which is now O(select.options.length + value.length)

We'll see what perf addicts say of this 👀

[GauBen]

Includes is fast! https://www.measurethat.net/Benchmarks/Show/6029/0/some-vs-filter-vs-indexof-vs-includes

[paoloricciuti]

It might be the better choice... I'm just saying that it seems an odd change I would like whoever wrote that code opinion on 😁

[paoloricciuti]

Why this change?

[GauBen]

It doesn't make much sense to have a templated function that does not use nor restrict its type parameter, it can be replaced with any. Asking for unknown[] makes it more type-safe, // @ts-ignore is no longer required.

[GauBen]

Thanks for the review @paoloricciuti!

For consistency, I'd say that <select multiple value={null / undefined}> should be accepted and ignored, like how other properties behave when given a nullish value. Regarding non-array values, TS can do the trick, we can hard-crash on them.