Conversation
🦋 Changeset detectedLatest commit: 8a962cd The changes in this PR will be included in the next version bump. This PR includes changesets to release 8 packages
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 |
teemingc
added
needs-platform-tests
This was referenced Mar 19, 2026
Member
teemingc
commented
Jun 4, 2026
teemingc
commented
Jun 4, 2026
Rich-Harris
removed
the
needs-ecosystem-ci
Member
|
Finally it looks like we have a clean run and can go back to opening PRs against this branch. Will see how much work is involved in fixing the conflicts on #15574 |
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
|
Install the latest version of pnpm add https://pkg.svelte.dev/@sveltejs/kit/c/8a962cd83ed2caf23de92ff9efd6ffef92531326Open in |
teemingc
added
needs-platform-tests
…error page (#16381) After an error page renders, `current.route` is `null` (`load_root_error_page` stores `route: null` in the navigation state). The next navigation computes ```js const route_changed = current.route ? route.id !== current.route.id : false; ``` so the route counts as unchanged no matter where the user goes next. A root layout server load that tracked `route` is then skipped as fresh, and its data from before the error page renders on the new page. The new test's failing assertion on `version-3` shows the layout still claiming the previous route: ``` Expected: "/b" Received: "/a" ``` A missing `current.route` now counts as changed. On a first-ever navigation the flag is irrelevant, the node has no previous data to reuse anyway, so the only behavior change is the conservative refetch after error pages. The interaction was never designed. `route: null` came in #6552 to give `beforeNavigate` a value meaning "no matched route", and the guard came two months later in #7450 with route tracking itself, mirroring the adjacent `url_changed` null-guard. Neither discussed the error-page case. The test lives in the no-ssr app. Its root layout gains a server load returning `route.id`, the navigation's data request is intercepted with a 500 to produce the error page, and the follow-up navigation must show the new route id. The app's root `+layout.server.js` was an empty file, shipped by #11354 so its 404-loop test has a server load node in dev; giving it a real load keeps that trigger intact. Found while reviewing #16380, which mirrors this computation for the error page itself and carries the same one-line fix. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed.
closes #16468 Warns but keeps working, same as the v2 `files` deprecation. The docs recommend `resolve.tsconfigPaths` per the issue; its friction with the generated tsconfig is tracked in #16471. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: Tee Ming <chewteeming01@gmail.com> Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com>
These were missing for the Netlify and static adapters
Standardises the Netlify adapter docs so that they're similar in structure to the other adapter docs
Supersedes #16428. It had the right approach, but we were actually handling `refresh` wrong in the first place. I thought we'd implemented this behavior but somehow we haven't. Anyway, this kills two birds with one stone: - It makes sure we allow `.refresh` (and `.set`) to run during render - It makes `.refresh` lazy: Calling `.refresh` busts the cache and guarantees that the cache will be populated by the end of the request. If you `await` the query or access its data, it will immediately run and populate the cache; otherwise SvelteKit will do it at the end of the request, in parallel with any other refreshing queries
closes #14315 This PR adds a utility for checking the spelling of hook files, checks the spelling for the universal hook file, and checks if an s is missing from "hooks". Also, only bothers with the check if the file itself has not been created. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] 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 - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed.
The `.wrangler` directory is generated by Cloudflare Wrangler when running the Wrangler worker preview and is placed relative to the wrangler config. This should be ignored no matter where it is, not just the root of the workspace
Prevents queries from infinitely looping during the drain phase. This has the side effect of preventing multiple refreshes of the same query during the drain phase, so... don't do that
…lds.foo.as(...) (#16331) closes #16321 Made the form ID a suffix so it's easier to read in the browser's inspector. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: Rich Harris <rich.harris@vercel.com>
The Vercel server handler files are in a `files` directory. This directory is where bundled output goes into for all the other adapter packages except for the Vercel adapter. As a result, it was accidentally included in the eslint ignore list. This PR changes that so it's no longer ignored (alternatively, we should just rename it as `src/`?)
Follow-up to #16795 — this applies the same thinking to `$app/state`. It allows us to get rid of the awful `BROWSER` hack and colocate logic more sensibly, with less indirection. Again, no changeset, since no user-observable changes. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com> Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com>
closes #16785 Basically, paths in the manifest are relative to Vite's configured `root`, but `write_client_manifest` was resolving them relative to `process.cwd()`. These are normally the same, but aren't when Vite is started programmatically from another directory, causing the generated client manifest to import route files from the wrong place. Passes the Vite root through to `write_client_manifest` and uses it to resolve components, universal modules, and param matchers before generating their imports. Adds a regression test where the project root differs from `process.cwd()`.
#16794) On Node, `setResponse` streams every `response.body` through a reader loop, so a response without a pre-computed `content-length` header is served with chunked transfer encoding. That is why the runtime can't move off the deprecated `json`/`text` helpers, whose only load-bearing job is stamping that header, and why a plain `Response.json` returned from an endpoint is served chunked today. Every other runtime derives the header from fixed bodies at serve time. `setResponse` now races the body's first reads against a `setImmediate` deadline before writing the head. A fixed body settles value-then-done on microtasks, so it is sent with a measured `content-length`; a body that leaves a read pending is a genuine stream and goes through the existing streaming path unchanged, with its headers delayed by at most one tick. Nothing is added when the response already carries a `content-length` or a `transfer-encoding` (proxied responses can carry the latter, and the pair would be invalid). Same approach as hono's node-server. `json`/`text` are untouched. Retiring them internally becomes a follow-up; its one catch is the `http.response.body.size` span attribute, which reads the `content-length` header. #16804 is stacked on this.
as suggested in #16754, modernizes the cloudflare adapter's docs to mention `wrangler types`. No changeset because docs-only --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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:`.
#16808 was merged too fast :(
Ref #15212. Something we've wanted to do for a while is get rid of the virtual modules, in favour of writing stuff to disk. The system becomes a lot easier to understand when it involves real artifacts instead of the crazy indirection we have going on at the moment. I started with the `$app/env/*` stuff because that's likely to be the most challenging, since it involves some mad science around starting up a mini Vite dev server to load the `src/env.ts` module so that we can analyse it so that we can create a generated module that _also_ loads the `src/env.ts` module... anyway, it works, and so I assume we will be able to do the same for the other `__sveltekit/*` modules. (We could probably replace the `resolveId` logic with an alias that just points to the `generated` folder, same as we have for `$app/*` currently. That can wait for a follow-up PR though.) Another thing I'm doing in this PR is creating the plugin in a separate module, rather than adding to the chaos in `vite/index.js`. It involves a little bit of duplication, but it makes everything so much more self-contained, and makes the coupling between different plugins more explicit (e.g. the `callback`). --- ### 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 - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com> Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com> Co-authored-by: Nic Polumeyv <nicolas.polum@outlook.com> Co-authored-by: Nic Polumeyv <162764842+Nic-Polumeyv@users.noreply.github.com>
closes #<!-- Add the related issue number here. Repeat this line for each additional issue it closes --> <!-- Explain the goal of the PR, why it is needed, and what has been changed to achieve that goal --> --- ### 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.
…ercel (#16809) closes #16806 In a pnpm workspace, `@vercel/nft` can trace the same dependency through two different symlink chains (e.g. `apps/app/node_modules/X` and `packages/ui/node_modules/X`) that both realpath to the same file in the pnpm store. `create_function_bundle` assumed traced file → destination is 1:1, so the second occurrence hit `fs.symlinkSync` on a `dest` that already existed and crashed the build with `EEXIST`. Wraps the symlink in a try/catch that ignores `EEXIST` — both chains would produce an equivalent link, so skipping the duplicate is safe.
`render_response` types `rendered` with a hand-written `{ head, body,
hashes: { script: string[] } }` and a TODO asking `svelte/server` to
expose `RenderOutput`. The file already imports `render`, so the type is
reachable today, and the hand-written version had drifted from it
(`hashes.script` is `` `sha256-${string}`[] ``).
`Omit` because `SyncRenderOutput` still carries the deprecated `html`,
which neither assignment site sets.
---------
Co-authored-by: Rich Harris <rich.harris@vercel.com>
) With #16794, `setResponse` derives `content-length` from fixed bodies, so runtime responses no longer need the `json` helper to be served correctly on Node. This switches the runtime's `json` call sites to the native `Response.json` (deprecated by #15448). `text` call sites stay: `new Response` with a string body adds a default `text/plain` content-type where the helper adds none, so replacing those changes response headers and is a separate decision. Tracing is unaffected, since page and error documents still go through `text`. Stacked on #16794; retarget to `version-3` once it merges. --------- Co-authored-by: Rich Harris <richard.a.harris@gmail.com>
…anonical form (#16339) closes #14781 During SSR, fetch responses are embedded as `<script data-sveltekit-fetched data-url=...>` tags, and during hydration the client rebuilds the url to find them instead of hitting the network. The two sides build it differently. The server serializes the normalized href from `new URL(input, event.url)`, while the client keeps the raw string, because prerendered pages may be served from any origin. That reasoning holds for same-origin urls, which are serialized path-relative, but cross-origin urls are absolute on both sides, so a raw string like `http://localhost:8080` without the trailing slash misses the cache and the request fires a second time from the browser. The fix normalizes `requested` to `resolved.href` in the cross-origin case only. Normalizing in `build_selector` instead, as suggested in the issue, would break the same-origin case, `new URL('/mock')` throws without a base. `requested` also became a `const` derived from `resolved` since the mutation was no longer needed. The new test fails on main, rendering `count: 2` after the hydration refetch overwrites the SSR data. Not fixed here: request bodies are hashed as strings on the server but dropped from the client hash when they are not strings, so POST-with-body fetches still miss the cache. Can follow up separately. Caching under the normalized href also meant a non-GET fetch spelled without the trailing slash could no longer evict the entry, so the second commit derives the eviction key the same way. Also not fixed: on prerendered pages the server classifies same-origin against `prerender.origin` while the client uses the live origin, so an absolute url pointing at the deploy origin still misses the cache. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed.
Follow-up to #16807. Instead of using a `resolveId` hook, we can use an alias for all the generated modules — every module ID like `<sveltekit:generated>/foo.js` corresponds to `.svelte-kit/generated/(build|dev)/foo.js`, making things a little easier to navigate, and reducing the cost of adding more generated modules relative to having to faff about with plugin hooks. The `<sveltekit:generated>` prefix is bikesheddable, but I figured it's worth being explicit about what this is, and using characters that are invalid in npm package names. Creating separate directories for dev and build means we don't need to be as careful about what goes where, and can freely use relative imports between generated modules. It means that building while also running a dev server won't result in clobbering. We can easily extend this to the other virtual modules. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com> Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com> Co-authored-by: Nic Polumeyv <nicolas.polum@outlook.com> Co-authored-by: Nic Polumeyv <162764842+Nic-Polumeyv@users.noreply.github.com>
…ng (#16838) #16449 made shallow `goto` run navigation hooks by copying two blocks out of `navigate` into `update_state`: the `onNavigate` callback run with its `afterNavigate` registration, and the completion tail (abort check, scroll and focus reset, `fulfil`, `to.scroll`, `afterNavigate` callbacks). #16685 then had to change both copies. They are now `run_on_navigate_callbacks` and `finish_navigation`, called from both places. The second commit builds `navigate`'s not-found fallback once instead of in two branches that differ only by the message suffix.
…the file's extension (#16837) `create_manifest_data` warns `Did you mean +${file.name}?` for any file whose name matches either the component or the module grammar, ignoring the extension. For `routes/a/error.ts` it suggests `+error.ts`, and renaming to that throws `Files prefixed with + are reserved`. The two grammars were also written out three times in the file; they're now two module-level patterns and the typo check picks the one matching the extension. Co-authored-by: Nic Polumeyv <nicolas.polum@gmail.com>
…ge to the correct trailing slash (#16836) The trailing-slash redirect for prerendered pages in `exports/vite/preview/index.js` writes an absolute `location` from the pathname the middleware sees, but that middleware runs inside `scoped(base, ...)`, which has already stripped `paths.base`. With a base path the browser lands on `/nested/` instead of `/base/nested/`. Production (#9351) and adapter-node (#16431) send a relative location for this reason and use 308; preview now does the same via `relative_pathname`. --------- Co-authored-by: Nic Polumeyv <nicolas.polum@gmail.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> Co-authored-by: Tee Ming <chewteeming01@gmail.com>
This PR gets rid of all the virtual modules, in favour of using real modules in `.svelte-kit/generated`. --- ### 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 - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com> Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com> Co-authored-by: Tee Ming <chewteeming01@gmail.com>
Member
Author
|
The release action is broken because of the number of commits on the branch changesets/changesets#2247 EDIT: it somehow fixed itself? |
…on-existent routes (#16376) closes #12910 When the client renders an error page for a URL that matches no route, the root layout's server data goes missing. The universal `+layout.ts` load receives `data: null` even though the server runs that exact load function to render the HTML 404. Reported with `ssr = false`, but any client-side navigation to a non-existent route hits it. ``` GET /this-route-does-not-exist/__data.json?x-sveltekit-invalidated=1 before 404 text/html (the rendered error page) after 200 application/json {"type":"data","nodes":[{"type":"data","data":[{"rootlayout":1},"rootlayout"],"uses":{}}]} ``` `load_root_error_page` fetches `__data.json` when the root layout has a server load, sending `x-sveltekit-invalidated=1` for the single root node. The server matches no route and falls through to the `state.depth === 0` branch in `respond.js`, which renders the HTML error page. That branch predates data requests, so they were never special-cased there, and the render runs the root layout server load only to discard the result as HTML. Client-side `load_data` then throws `HttpError(404)`, which since #16135 is deliberately swallowed to avoid a reload loop, leaving `server_data_node` null. teemingc diagnosed this on the issue and named two acceptable outcomes, "We need to be able to return layout data for an error page although a route doesn't exist or just return the fallback error page." This PR implements the first, server side only. The client already consumes the response, and the fallback option would discard data the server computes anyway while rendering the HTML 404. No client changes. The new branch only fires for `is_data_request && invalidated_data_nodes?.length === 1`. The invalidation parameter marks kit's own `load_data` fetches, so a browser navigating directly to a `.../__data.json` URL keeps getting the HTML error page that #15884 deliberately improved (its options-2 test still passes). Length 1 is exactly the shape `load_root_error_page` sends, so an old deployed client requesting multi-node data for a route that no longer exists on the new server keeps today's 404 instead of receiving a truncated nodes array. Prerendering is excluded so a missing path still reports 404 for dead-link detection. The response reuses `render_data` with a synthetic root page (`layouts: [], leaf: 0`), so load execution, serialization, streaming, redirect and error nodes all behave as they do for matched routes. Its `route` JSDoc is narrowed to the two fields it reads instead of casting the synthetic object to a full `SSRRoute`. Static hosts are unaffected, there is no server to return the data, and #16135 already keeps the fallback page from looping there. The new test in the basics `Errors` describe asserts both halves, JSON with the parameter and HTML 404 without it. Fails on `version-3` without the `respond.js` change. Sibling PR #16380 touches the client half of this flow (reusing fresh root layout data instead of refetching). The two compose, whichever lands second rebases trivially. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: Nic Polumeyv <nicolas.polum@gmail.com>
…non-ActionResult error response (#16308) closes #15737 Submitting a `use:enhance` form that trips the CSRF origin check does nothing visible. The 403 response is right there in the network tab: ```json { "message": "Cross-site POST form submissions are forbidden" } ``` but it has no `type`, so it isn't an ActionResult and every branch in the submit handler and `applyAction` skips it. Non-JSON responses already become `{ type: 'error' }` through the catch around `deserialize`, so JSON that isn't an ActionResult was the one shape that failed silently. Error responses without a recognized `type` now throw into that same catch and render the nearest `+error.svelte`. A body shaped like an `App.Error` becomes `page.error` as-is, the way an `error(403, { message })` body does. Anything else goes through `handleError`, which #16162 routed this catch through, so the hook keeps seeing these failures and `page.error` keeps its declared shape. 2xx responses are untouched. PatrickG suggested rendering the error page in the issue. teemingc flagged the same gap in #10464 with a server-side shape fix in mind; doing it on the client also covers proxy and middleware responses that kit's server never shaped. #10855 reports the same class of unhelpful failure for non-action endpoints; the non-2xx half of it is covered here. Responses that do parse as an ActionResult pass through regardless of status, which keeps the pattern that prompted the #13197 revert (#13397) working. The docs line that revert added says posting to a `+server.js` endpoint results in an error; with this change that error surfaces instead of failing silently. The test mimics the CSRF response with an endpoint, since the real check can't fire same-origin in Playwright. It fails on `version-3` and passes with this change, in dev and build. The hook suffix in two of the assertions is `handleError` running. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: Nic Polumeyv <nicolas.polum@gmail.com> Co-authored-by: Rich Harris <rich.harris@vercel.com>
…16443) Implements the `name` property from #16424 (comment). `name` is looked up from the Vite client manifest where the fonts arrays are built, so the filter receives the exact source file name. Deriving it from the emitted path was wrong two ways, the bundler sanitizes characters (`inter+bold.woff2` is emitted as `inter_bold.<hash>.woff2`) and content-identical assets are emitted once under a single name. The union includes `name` for `font` only, kit never calls `preload` with type `asset`. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: Rich Harris <rich.harris@vercel.com>
`read_stream` accumulates decoded text in one string, re-runs `indexOf` over the whole buffer after every transport chunk and re-slices the buffer for every emitted record. When one frame spans many chunks, every prefix is rescanned and total work grows quadratically with frame size. This parser sits behind streamed page data and remote functions (NDJSON) and `query.live` (SSE), so the cost lands on the client for large payloads. With this change only newly decoded text is searched. Already searched text accumulates in an array and is joined once per completed frame. The last `delimiter.length - 1` characters stay in the unsearched tail so a delimiter split across chunk boundaries still matches. One 2 MiB frame, Node 22 x64, median of repeated runs: | transport chunks | before | after | speedup | | ---: | ---: | ---: | ---: | | 2048 × 1 KiB | 1720.7 ms | 4.0 ms | 429× | | 512 × 4 KiB | 431.0 ms | 3.2 ms | 135× | | 128 × 16 KiB | 111.5 ms | 3.1 ms | 36× | | 32 × 64 KiB | 31.3 ms | 3.1 ms | 10× | Many small records, the common path, get slightly faster (2 MiB of 64 byte records in 16 KiB chunks, ~21 ms to ~15 ms) because the old code re-sliced the buffer once per record. Output is unchanged. The rewrite matched the previous implementation across 4000 randomized cases covering both delimiters, delimiters split between chunks, multibyte UTF-8 split between chunks and trailing unterminated records. The new unit tests also pass against the previous implementation. `read_stream` was extracted in #15957 and last changed in #16423. No open PR modifies `stream.js`, `ndjson.js` or `sse.js`. --- ### 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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:`.
…ch parameter changes (#16495) Since #11258, `diff_search_params` compares each key's values as a set, so navigating from `?x=a&x=a` to `?x=a` is treated as unchanged and load functions tracking `x` don't rerun. Comparing the sorted value lists instead catches count changes, still treats reordering as unchanged, and is O(V log V) rather than O(V²). ### 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 2.0 (major releases only that fix regressions) ### Edits - [ ] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed.
`core/env.js` had one generator per `<sveltekit:generated>/env/*` module, each iterating the same variables and calling `validate()` on the same static values, so a user's schema ran three to four times per build (and per hot update in dev), and a validator that isn't a pure function of its input baked different values into `config.js`, `public/server.js` and `public/client.js`. `create_env_modules` replaces them with a single pass over the variables that returns every module keyed by path, so the plugin only writes files. Dynamic variables are untouched; they were already validated once by `set_env`. One visible change: invalid variables that previously surfaced one module at a time are now reported in a single error.
`client.js` walks `query_map` and `live_query_map` eight times across `_invalidate` and `_goto`, each as a hand-written `for (const [id, entries] of map) for (const [payload, entry] of entries)` pair, and five of those recompute `create_remote_key(id, payload)` inline to get the key the consumer actually wants. This adds a `cache_entries(map)` generator yielding `[key, entry]` and collapses each site to a single loop. No behaviour change; the walk order and the set of resources touched are identical. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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:`.
`handle_action_json_request` in `runtime/server/page/actions.js` was a copy of `handle_action_request` that differed only in how each branch was encoded, and the 405 result was written out a third time in `handle_remote_form_post_internal` (`runtime/server/remote-functions.js`). #16684 had to add `location` to every copy. `handle_action_request` now runs once and `action_result_json` encodes the result for the JSON transport; `method_not_allowed_result` and `action_error_result` are shared with the remote form POST handler. The `fail()` misuse check is no longer in the shared catch, since `fail()` is form-action only. Remote `form` handlers get a dev-time error for a thrown or returned `fail()` in `app/server/remote/form.js`, covering both the enhanced and no-JS transports. `ServerActionResult` (`types/internal.d.ts`) types the pre-`handleError` result, so we're no longer lying about the types. --------- Co-authored-by: Rich Harris <rich.harris@vercel.com>
…6720) closes #16696 `validate()` checks `element` before `await tick()`, but the attachment cleanup sets `element = null` when the `<form>` unmounts. If the form is removed while `validate()` is waiting for that tick — e.g. an `{#if}` flips right after a fast submission — the resumed call throws: ``` TypeError: Cannot read properties of null (reading 'querySelector') ``` The fix re-checks `element` after the tick and bails out if the form is gone, which matches what the pre-tick check already intends. ### Tests Added a case to the `async` test app's existing `/remote/form/validate` route: a form inside an `{#if}` block, and a button that starts `validate()` and unmounts the form in the same handler, catching whatever the promise rejects with. The new test fails on `version-3` (`TypeError: Cannot read properties of null (reading 'querySelector')`) and passes with this change. Rebased onto `version-3` at a115a7b and re-verified locally on Node 24.18.0: - `pnpm lint`: 0 errors (one pre-existing warning in `src/runtime/server/page/index.js`, untouched here). - `pnpm test:unit`: 895 passed, 11 failed. The 11 failures are in `kit-prerendering-paths-base` and `kit-prerendering-options` and reproduce identically on unpatched `version-3`, so they are not from this change. - `async` test app, `pnpm test:dev`: 193 passed, 103 skipped, 0 failed. The new test also passes in `chromium-build` and `chromium-build-no-js`. - `pnpm check` fails on unpatched `version-3` as well, with 24 `TS2300: Duplicate identifier` errors from `src/types/ambient.d.ts`. This branch changes no types, so I could not get a clean baseline for that step. --- ### Please don't delete this checklist! Before submitting the PR, please make sure you do the following: - [x] 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 - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. ### Tests - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` and `pnpm check` ### Changesets - [x] 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 - [x] Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed. --------- Co-authored-by: Elliott Johnson <hello@ell.iott.dev> Co-authored-by: Rich Harris <rich.harris@vercel.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Opened just so that I could easily see CI checks
You can try SvelteKit 3 by installing any of the packages following #15497 (comment) or the ones published with the
nexttag https://npmx.dev/package/@sveltejs/kitPlease don't delete this checklist! Before submitting the PR, please make sure you do the following:
Tests
pnpm testand lint the project withpnpm lintandpnpm checkChangesets
pnpm changesetand following the prompts. Changesets that add features should beminorand those that fix bugs should bepatch. Please prefix changeset messages withfeat:,fix:, orchore:.Edits