fix: catch error on @const tag in svelte:boundary in DEV mode

[adiguba]

• Close svelte:boundary doesn't catch errors inside a @const tag #15368

I'm not sure to understand the reason, but in DEV mode, the @const tag add a $.get() "to eagerly evaluate the expression in order to hit any 'Cannot access x before initialization' errors"

When used in a <svelte:boundary>, this code need to be inside the $.boundary().
So I added const_dev_statements property into SvelteBoundary which is populated by ConstTag when his parent is a Svelteboundary.

Not sure if this need a test, neither how to test in dev mode.

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.
• 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: 0ab422b

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

[svelte-docs-bot]

Preview: https://svelte-dev-git-preview-svelte-15369-svelte.vercel.app/

[paoloricciuti]

We definitely need a test for this

neither how to test in dev mode.

you can specify compileOption.dev = true in the config

[adiguba]

Test added.
But I'm not sure how to handle the conflict on index.d.ts...

[paoloricciuti]

But I'm not sure how to handle the conflict on index.d.ts...

Should be fixed now

[github-actions]

Playground

pnpm add https://pkg.pr.new/svelte@15369

[adiguba]

Please wait to merge.
I dislike the way I handle that.
I think I can do it more simply...

[adiguba]

I think it's way better like that, since it modifies fewer files...

[paoloricciuti]

I think it's way better like that, since it modifies fewer files...

I agree...much better now...let me just check a couple of things as soon as we get the actual playground

[paoloricciuti]

I was trying to understand if the dev flag was needed but while exploring this I think the whole accessing eagerly const tags is not really needed. Commenting the line make no test fail... @dummdidumm you added that line here ...any insight?

[trueadm]

Thank you

[trueadm]

I was trying to understand if the dev flag was needed but while exploring this I think the whole accessing eagerly const tags is not really needed. Commenting the line make no test fail... @dummdidumm you added that line here ...any insight?

It's useful because deriveds are lazy, and by reading it eagerly in DEV to capture any Cannot access x before initialization errors. I believe we even have a test for this from Svelte 4?

[paoloricciuti]

It's useful because deriveds are lazy, and by reading it eagerly in DEV to capture any Cannot access x before initialization errors. I believe we even have a test for this from Svelte 4?

Removing that line was not failing any test...how can you have a cannot access x before initialization in a const tag?

[trueadm]

@paoloricciuti You can reference something in one const tag vs the other and it can cause that issue.

[paoloricciuti]

@paoloricciuti You can reference something in one const tag vs the other and it can cause that issue.

Const tags are reorganized to not have the issue and if there's some cycle it's a compiler error

[trueadm]

@paoloricciuti You can reference something in one const tag vs the other and it can cause that issue.

Const tags are reorganized to not have the issue and if there's some cycle it's a compiler error

I believe that's not the case for where you reference things like snippets right?

[paoloricciuti]

I believe that's not the case for where you reference things like snippets right?

Oh that's it