fix: repair href attribute mismatches

[dummdidumm]

The reasoning in #9662 turns out to be false: While it's true that the href would resolve to the same URL on the initial page, once you're doing a client-side navigation to a URL at a different depth (and the link stays around, for example because it's in a layout), the relative path would now point to the wrong location. Therefore we need to repair href hydration mismatches

fixes sveltejs/kit#12254

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: fbd234a

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]

I'm not following — if you're navigating client-side then you're not hydrating, so you wouldn't reach this code

[dummdidumm]

Imagine you have a layout like

<!-- +layout.svelte -->
<script>
  import { base } from "$app/paths";
</script>

<a href="{base}/">home</a>
<a href="{base}/foo">foo</a>

<slot />

You start at /base-path/, so the href for foo looks like ./foo. Now you click on that link, so you're on /base-path/foo. Since this is the root layout, it will stay as-is between navigations, so the href still links to ./foo. That means, when you click on it again, you end up on /base-path/foo/foo -> wrong.

[Rich-Harris]

Ah, I see. We still need to ignore mismatches for <link> elements, otherwise changing ./foo to /foo will trigger another network request (the avoidance of which was the purpose of ignoring href mismatches in the first place, just like for src and srcset), so we should reinstate the test but change the <a> to a <link>.

The other cases where an href attribute applies are <base> and <area>, neither of which eagerly trigger requests, so it's just <link> that this applies to AFAICT.