fix: work around safari bug

[Rich-Harris]

Ref #11364. There is a very serious Safari bug that apparently very few people have encountered outside SvelteKit, possibly because few frameworks use native ESM with top-level await. That is likely to change over time, and I hope Webkit can prioritise a fix for this issue.

In the meantime we have no choice but to work around it. I don't see a better alternative than this PR, which eagerly loads the dynamic env.js module when on a prerendered page, such that ${global}.env is populated by the time any user code tries to read it.

This is less good than fetching it lazily, and it only mitigates the issue (since top-level await could appear in user code, and it's likely we'll find other uses for it within the framework in future), but for now I think it will have to do.

---

Please don't delete this checklist! 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
• This message body should clearly illustrate what problems it solves.
• Ideally, include a test that fails without this PR but passes with it.

Tests

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

Changesets

• If your PR makes a change that should be noted in one or more packages' changelogs, generate a changeset by running pnpm changeset and following the prompts. Changesets that add features should be minor and those that fix bugs should be patch. Please prefix changeset messages with feat:, fix:, or chore:.

Edits

• Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed.

[changeset-bot]

🦋 Changeset detected

Latest commit: 4e84a12

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

This PR includes changesets to release 1 package

Name	Type
@sveltejs/kit	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]

Not sure why the tests are failing here, they succeed locally

[Rich-Harris]

Gah, it fails when $env/dynamic/public is imported in hooks.client.js, because the module is evaluated before ${global}.env can be populated.

This might mean we need to wait for env.js to be imported before importing the actual app, which would really suck.

[bfanger]

Does code in the client.app (or start) use the values of env during the import phase? Because that could be affected with this change.
(Thats why i load the env before everything else in my fix attempt https://github.com/bfanger/kit/tree/fix/sveltekit-v2-safari)

[Rich-Harris]

start doesn't, but app could because hooks are imported eagerly, and your hooks.client.js could import $env/dynamic/public.

So our solutions are essentially the same — load env, then load everything else to ensure that evaluation happens in the correct order. Because of preloading it shouldn't really make any observable difference but it's still kinda filthy.