fix server crash when accessing a malformed URI

[ivanhofer]

Fix for #5090

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. All changesets should be patch until SvelteKit 1.0

[changeset-bot]

🦋 Changeset detected

Latest commit: be026cd

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

[ivanhofer]

I'm not sure how to fix the test. I would need some help there.

[benmccann]

@ivanhofer I'm not sure if you've noticed, but the tests appear to be failing with this PR

[ivanhofer]

@ivanhofer I'm not sure if you've noticed, but the tests appear to be failing with this PR

Yes I saw. Thats why I left the comment above. I would expect the test to work that I have written. In CI two tests are failing. Locally only the test with javaScriptEnabled fails because it waits for something that is not present on the page.
And I don't really understand how I can make the test not fail. Any hints?

[benmccann]

Ah, sorry, I didn't see your comment.

It seems like the failure is coming from Vite. Probably the best course of action would be to create a small project that reproduces this using only Vite and not SvelteKit and file a bug there. See https://github.com/sveltejs/kit#bug-reporting for more details on that

It's not clear to me that this PR actually fixes anything. Is there a reason you think it might? Or should we go ahead and close this? At the very least I would mark it as a draft to make it clearer that it's not in a state to be merged

[ivanhofer]

I don't think this has anything to do with vite. It is an error that SvelteKit does not handle yet. I think this might only occur when using the node adapter. I just added this info to the issue description.

The fix works. We are using it since a few weeks in two of our applications.
Also when testing it locally I get the expected outcome : a 400 error.
The tests in this CI process deliver a 500 error and I don't really get why. That's why I'm asking for help because I could not figure out what happens inside this test setup and I don't want to change something that could affect other tests.

[benmccann]

The 500 on the CI is coming from Vite according to the stack trace. Did you test this fix with your own app in both dev and build mode? The Vite issue is only occuring in dev mode since that's where the Vite dev server is used

[ivanhofer]

Ah, now I see that there are two different "test" calls "test:dev" and "test:prod". So there is actually a bug in SvelteKit and in vite. I'll also open an issue there.

Still there is an error when running "test:prod" with the javaScriptEnabled option, that does not come from vite. It comes from how the tests are setup.

[benmccann]

I took a look at filing a Vite issue this morning, but I think Vite's mostly fine because the server doesn't crash entirely like the SvelteKit server does (although it could arguably return a 400 status code instead of 500). I think we just need to update the test in this PR

[ivanhofer]

Smart way to handle the test 👍
Thanks for your investigation and help!

[Rich-Harris]

thanks!