[pull] canary from vercel:canary #122

Follow up from #69695.

…ment (#69691) While investigating #65614 I found that in Next.js with webpack we mark this dependency as external but in Turbopack we didn't do that yet, causing it to be bundled and then fail at runtime because the path didn't match exactly. This PR ensures it's marked as external. Didn't add an additional test as we still have some Turbopack build test failures related to opengraph images so potentially this change also fixes those 💯 Fixes PACK-3058 Fixes #65614

Similar to #68969, skips a few more tests that are checking features that we don't have to report on.

### What? Pass `resourceQuery` to loaders correctly. ### Why? To be consistent with webpack ### How? Closes #69502 --------- Co-authored-by: Tobias Koppers <tobias.koppers@googlemail.com>

Following up from: #69614 Add missing switcher prop to code blocks, and fix page formatting.

## What? Implements `--experimental-app-only` for Turbopack build.

Co-authored-by: Jiachi Liu <inbox@huozhi.im>

Found a few more tests that are related to custom webpack configuration.

### What? * improve error * fix CachedTaskType serialization * fix next config serialization * fix serialization of Rope * fix serialization of AliasMap * more efficent rope serialization * add missing primitives * Disable serialization for RuntimeVersions

add TODO fix function name

Fix typo: `prender` → `prerender` Co-authored-by: JJ Kasper <jj@jjsweb.site>

Co-authored-by: Jiwon Choi <devjiwonchoi@gmail.com>

```diff - unmanged + unmanaged ``` Is annoying when searching for `unmanaged` in the file.

This adds new information to the routes manifests that is used during deployment for pages rendered using Partial Prerendering (PPR). It changes the route detection mode from using a pathname to detect a PPR resume request to using a new header, `Next-Resume: 1`. When a `POST` request is received with this header, it is assumed to be a resume request. The old mechanism is being retained until a future canary release to validate the changes to infrastructure.

Some basic refactors that I made while browsing the fonts code. No behavioral differences should be observed so existing tests suffice.

Removes the in-between test that relies on the order of test execution which is not needed as this can throw in beforeAll to make all relevant tests fail.

MInor typo  Co-authored-by: JJ Kasper <jj@jjsweb.site>

When another data request is received by the pages router, a cancellation error is thrown and is meant to be caught by `change` to signal that the navigation request was aborted. However we were logging the error right before we handled the logic to properly bubble the error up. This meant that it was logging an exception to the console that wasn't actionable. Fixes #40554 Closes NEXT-3731

Since this is in `test/integration` the path with Turbopack will start at the repository root, it's still correct.

… helper (#69762) Noticed while reading through this code. [As of Rust 1.77.0, recursive async functions are allowed inline, as long as they use indirection](https://blog.rust-lang.org/2024/03/21/Rust-1.77.0.html#support-for-recursion-in-async-fn). This cleans up the helper method.

### What? State need some special handling for persistent caching. In particular we need to differ between two kind of states: * Persistent State: It will be stored in the persistent cache and value will be kept between builds. * Transient State: It's only valid for a session and will reset when restoring the persistent cache. We didn't have the separating before, so this PR introduces a new type `TransientState<T>` next to `State<T>` which handles transient state. The value will always be an `Option<T>` and resets to `None` when restoring the persistent cache. This also means all task that depend on transient state will be invalidated on restoring. Transient State can also be used when the value is not serializable. e. g. this is the case for the `last_successful_parse` state in ecmascript modules. We use that to avoid large structure changes to the module graph when introducing parse errors to modules. Using transient state for that means we only apply this optimization on a session, and it resets when restoring from persistent cache.

### What? evaluation require.resolve tries to resolve and process a module during module analytics. Since analytics is part of processing a module this can create a call cycle when `require.resolve("./self")` is used. This happens in some npm packages and caused Turbopack to hange. This PR fixes that problem by only resolving the module and not processing it when using `require.resolve`. In addition to that it adds a panic when a call cycle occurs. The next time we mess it up and create a call cycle this will crash Turbopack instead of hanging, which makes it much easier to find the problem. fixes PACK-3227

Fixes two tests for tsconfig/jsconfig being empty.

This ensures we properly set the request meta during our middleware invoke so that the `NoFallbackError` bubbling is handled properly. Without these fields set we bubble the error unexpectedly causing an unexpected 500 instead of a 404. x-ref: #66987 Closes: #69428

@lubieowoce

…9806) There's code that uses `glob` to find all template files and replaces the default import alias with whatever is specified during CNA. This does so without excluding fonts, and so we're unintentionally corrupting these `woff` files. This wasn't a problem in previous versions because we didn't use `localFont` in the default template, just `Inter`. The files were technically still being corrupted it just never manifested unless you went to actually use them. This is a quick fix to introduce minimal changes but ideally in a follow-up we should figure out a better solution for replacing aliases, ie by using an allowlist rather than an exclude list. h/t to @lubieowoce for the thorough test cases Fixes #69748 --------- Co-authored-by: Janka Uryga <lolzatu2@gmail.com>

## Why? A manual workflow that will query issues by version string (e.g., `14.2.8`), then output the relevant issues to Slack. x-ref: https://vercel.slack.com/archives/C04LGE5SYEB/p1725659202909009

This test assumed a hardcoded file but with Turbopack it's a combination of multiple files. Changed it to use the manifest to look up the relevant files.

…ase (#69804) ## What? Fix `find_turbopack_part_id_in_asserts` , which is not looking for "export ..." strings so it returns `None` in cases where it should return `Some(PartId::Export(e))`. ## Why? This is making the `part_of_module` function not behave exactly as expected, as it [creates a `PartId::Export`](https://github.com/vercel/next.js//blob/09-06-fix_find_turbopack_part_id_in_asserts/turbopack/crates/turbopack-ecmascript/src/tree_shake/mod.rs#L655) passed as `NamedExport.with` and [later processed with `parse_with`](https://github.com/vercel/next.js//blob/09-06-fix_find_turbopack_part_id_in_asserts/turbopack/crates/turbopack-ecmascript/src/analyzer/imports.rs#L369) during the analysis, which calls `find_turbopack_part_id_in_asserts`.

When returning an aliased prefetch, we need to make sure the final tree that is produced contains the final searchParams, as otherwise the router might not properly re-key the component to reflect the changed search value. This moves the existing util that handled appending `searchParams` to the `__PAGE__` segment into the shared segment helpers. On the client, when we detect that we returned an aliased prefetch, we update the segment key with the "final" data by calling this helper with the final searchParams. This also fixes a missing `aliased` detection in the case where we do find an entry as it's possible the entry matches the path, but not the search. It wasn’t sufficient to only do this for full prefetches because while auto prefetches won’t return the CacheNode data, it will return a tree, so we need to make sure the tree contains the right params. Fixes #69847

@arlyon

…turbopackIgnore (revision of #69113) (#69768) ## History @arlyon was working on this in #69113. He's out of office, and this is a feature we want to get landed soon, so I was seeing what needed to happen to get it to merge. I encountered more issues, and I didn't feel entirely comfortable pushing deep changes to @arlyon's branch, so this builds on top of his (rebased) commit. This is my attempted PR to fix these problems. ## What? *Copied from #69113's description.* Certain bundles may be designed specifically to be included at runtime. To support this usecase, webpack implemented a comment directive `webpackIgnore` that opted out that particular import from bundle time optimisations. We need to support this, as some libraries take advantage of this (mapbox). Closes PACK-3046 ## Changes from #69113 - Bail out much earlier in the analysis. The other PR appeared to bail out too late, and would generate a `Promise.resolve()` call, instead of preserving `import`/`require`. My understanding of webpack's feature is that it should include the `import`/`require`. - Skip "too dynamic" analysis for ignored imports. - Switch to a snapshot test, so that we can verify that the output actually contains the `import`/`require` expressions. - Integrates my changes from #69755 to use an `ImportOverrides` struct instead of a bool. --------- Co-authored-by: Alexander Lyon <arlyon@me.com> Co-authored-by: Niklas Mischkulnig <4586894+mischnic@users.noreply.github.com>

### TL;DR Added support for a new `experimental.turbo.root` configuration option in Next.js. ### What changed? - Introduced a new `root` property in the `ExperimentalTurboOptions` interface. - Updated the build process and hot reloader to use the new `experimental.turbo.root` configuration. - Added fallback logic to use `outputFileTracingRoot` or the project directory if `experimental.turbo.root` is not set. - Implemented warning messages for non-absolute paths in `experimental.turbo.root` and `outputFileTracingRoot`. - Added a utility function `dset` for safely setting nested object properties. ### How to test? 1. Set up a Next.js project with Turbopack enabled. 2. Add the `experimental.turbo.root` option to your `next.config.js`: ```javascript module.exports = { experimental: { turbo: { root: '/path/to/your/root' } } } ``` 3. Run the development server and build process to ensure they use the new root path. 4. Verify that file resolution works correctly with the new root path. ### Why make this change? This change allows developers to explicitly set the root directory for Turbopack, providing more control over file resolution and project structure. It enhances flexibility for monorepo setups and projects with complex directory structures, ensuring that Turbopack can correctly resolve files above the project directory when necessary. Closes PACK-2646 Fixes #62409

## What? This implements passing the browerslist config from Next.js to Turbopack. Initially I implemented passing the user browserslist config in this PR as well but found that when this is lowered it causes test failures, those need to be investigated separately but shouldn't block these changes landing.

Step 1: detect `new Worker(new URL(..., import.meta))` calls Detect `new URL("..", import.meta.url)` and turn into `JsValue::Url`

Closes PACK-2504 Closes PACK-3020 - Implemented `WorkerAssetReference` to handle `new Worker()` calls - Added `WorkerLoaderModule` and `WorkerLoaderChunkItem` to create separate chunk groups for worker scripts - Updated runtime code to support loading worker scripts - Modified the analysis process to detect and handle worker creation - Workers are loaded via blob urls (which then `importScripts`-s all required chunks), which required introducing a `TURBOPACK_WORKER_LOCATION` global variable to support the relative chunk urls in the isolated blob origin. - [x] `evaluated_chunk_group` is not implemented by `NodeJsChunkingContext`, so it currently fails to build in RSC/SSR contexts - I've made it ignore worker calls for `Rendering::Server` for now

This PR refactors the Link component docs page to follow a consistent structure according to our reference docs template. - Adds both typescript and javascript examples to all code blocks - Adds bot pages and app router examples to all code blocks where applicable - Re-structures the page Fixes: https://linear.app/vercel/issue/DOC-3258/nextlink  --------- Co-authored-by: Delba de Oliveira <32464864+delbaoliveira@users.noreply.github.com> Co-authored-by: Delba de Oliveira <delbabrown@gmail.com>

### What? only subscribe to changes in dev mode ### Why? this avoid bundling twice due to data endpoint

If a Rust Error Result is returned by an hmr update, then an unexpected internal error unrelated to the user’s code has occurred. Like #67678 does for unhandled panics, log these errors to a file and inform the user with a generic message informing them to file an issue. In development, this continues to log to stderr.

…pers, and remove the async-recursion macro (#69805) An extension of #69762: > There are also a few more of these helpers if you search for `_boxed(`. These things used to be needed before Rust 1.77.0: https://blog.rust-lang.org/2024/03/21/Rust-1.77.0.html#support-for-recursion-in-async-fn These days, we can just do the boxing inline! I also noticed we could remove the `async-recursion` macro, as it's not providing much value on Rust >=1.77.0.

…g methods, improve guide (#69616) Closes: https://linear.app/vercel/issue/DOC-3297/draftmode-reference-page - Documents missing methods ([source](https://github.com/vercel/next.js/blob/adc94b49d47cd263140ba6523bc2db2e6d09e0d2/packages/next/src/client/components/draft-mode.ts#L6)) - Adds examples - Review and improve Draft Mode guide. Will add the steps component once we start working on the guides. --------- Co-authored-by: Rich Haines <hello@richardhaines.dev>

#69822) fixes #69820 - fixes wrong protocol issue when you run `next dev` with `--experimental-https` and `-H` option. To check the PR fixes the issue correctly, run the command below. ``` corepack pnpm next -H 0.0.0.0 --experimental-https YOUR_APP_DIRECTORY ```  Co-authored-by: JJ Kasper <jj@jjsweb.site>

### What? Remove `console.log(err)` in `isDynamicPostpone`, I think this should be a debug console that was accidentally left behind. ### Why? This will cause meaningless logs to appear in the console when we use the `unstable_rethrow` API. Co-authored-by: JJ Kasper <jj@jjsweb.site>

`__non_webpack_require__` is not supported in Turbopack (webpack specific)

@samcx

…69771) This PR updates the modularize-imports example to use the App Router. Here are the changes that have been made: - Renamed the "pages" folder to the "app" folder. - Updated the routing for / files to align with the App Router. - Added the layout.tsx file as part of the App Router. - Updated the package.json file. CC: @samcx --------- Co-authored-by: Sam Ko <sam@vercel.com>

This test assumed the code would be in the entry file, but that is not always the case, and with Turbopack we even have more chunks so the files can be in both `.next/server/pages` or `.next/server/chunks`. Updates the test to reflect that, replacing was already implemented in Turbopack.

This continues #39707 bringing the changes up to date with canary and adds test cases to ensure it's working as expected. Closes: #22319 Closes: #39707 Closes: NDX-148

## Why? The current docs on scrolling behavior when using `<Link>` are misleading and incorrect. This edit correctly documents the expected scrolling behavior (in both `true` or `false` cases). - https://nextjs.org/docs/app/api-reference/components/link#scroll

Pages router (`/pages`) will continue to support React 18 not the React 19 RC. Current thinking is that we'll add support for React 19 in Pages Router once 19 is stable. This does not affect App Router (`/app`) which continues to use the latest React Canary (i.e. React 19). #65058 is required reading to understand the changes in this PR --------- Co-authored-by: Zack Tanner <1939140+ztanner@users.noreply.github.com>

## Why? To prevent confusion about the `/public`, re-adding to Create Next App. x-ref: https://x.com/YTCodeAntonio/status/1832856105447473351

…App Router (#69896) ## Why? Adding onto this → #69894. A feel more additions to clarify the scroll behavior in App Router. ## How? - Remove instances of previous, incorrect way of describing the default scroll behavior - Remove **Disabling Scroll Restoration** section - Refactor **Disable scrolling to the top of the page** section

Previously, in a number of situations, we would implicitly use the entry module’s ident to name the chunk group. This adds the ability to specify a different name. Now, for evaluated chunk groups for pages, we use the page source’s ident. This prevents changes in loader chunk groups from changing the chunk name, which breaks HMR. We still use the entry module’s ident for most other cases now, but it is now explicitly provided. Test Plan: `TURBOPACK=1 pnpm test-dev test/development/acceptance/ReactRefreshModule.test.ts` with #68698 applied --------- Co-authored-by: Tobias Koppers <tobias.koppers@googlemail.com>

Reverts #69484

With Turbopack the output path is a bit different that's why the test couldn't find the `.css` file. Changed the test to ensure it checks for `.css` files existing.

## Why? There is no mention that Pages are rendered on demand in Development, regardless if you use this route segment config. ``` export const revalidate = 10 ```

### What? Add a linter for importing `ImageResponse` from `next/server`. ### Why? It's deprecated . ### How? Closes PACK-3237

Ensures these tests check all output files instead of only the entry file. Given that turbopack can do further chunk splitting.

Fixes the test, matched development.

### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide x-ref: #69894  Co-authored-by: JJ Kasper <jj@jjsweb.site>

These two are testing webpack config.

#69569) Reland of #69141 after it got reverted. Turns out there was a GT submit that got lost (along with the associated commit) somehow. --------- Co-authored-by: Niklas Mischkulnig <4586894+mischnic@users.noreply.github.com>

…mory backend (#69472) #69663 made improvements to the display logic, we don't need as much logic here.

…rbo-tasks, represent Empty as None (#69473) This is preparation for local tasks/cells in #69126. - Moves this into the `turbo-tasks` crate so that it can be used with local cells/outputs in #69126. - Remove the `Empty` state from the enum because the implementation in #69126 does not currently need that. This can still be represented using `None`.

…mory/src/task.rs (#69505) This code was a bit inconsistent with naming/syntax/etc between the different branches, and it was bothering me. No functional changes in this PR!

@sokra

Closes PACK-3117 Closes PACK-3235 - Also replace `typeof` of `module`, `exports`, `require` in ESM with the corresponding value - Also replace `typeof import.meta` with `object`. This is what we talked about, @sokra

Closes #69861 Closes PACK-3240 The sourcemaps of server actions were invalid, the listed source file was the file `AssetIdent` like `turbopack://[project]/bench/basic-app/app/actions.ts [app-client] (ecmascript, action)` instead of only the source file path. This broke the sourcemap mapping logic of the error overlay. --------- Co-authored-by: Tim Neutkens <tim@timneutkens.nl>

### What? This PR fixes a text formatting error where the text shown after the Next.js version is bigger than the version itself. <img width="768" alt="CleanShot 2024-09-06 at 19 41 08@2x" src="https://github.com/user-attachments/assets/1ebe9d4d-860d-4f1a-95c4-36bf67e7b87b"> Fixes #69813 Co-authored-by: Sam Ko <sam@vercel.com>

## Why? Updating the channel for popular issues. The updated channel will serve the action's purpose better than its current channel.

### What? Use https://vercel.com/docs/functions/runtimes/edge-runtime#compatible-node.js-modules for the source of truth. ### Why? x-ref: https://vercel.slack.com/archives/C03S8ED1DKM/p1725654625787459?thread_ts=1724702057.947199&cid=C03S8ED1DKM ### How?

…tream image is identical (#67257) ### What? - Reduces redundant image optimizations if the upstream image is unchanged, there is no need to run the optimization again as you can just use the previously optimized image. - Changes `getHash`-function to return `base64url` instead of custom variant of standard `base64`. This can be removed if deemed not necessary. - Moves all image ETag calculation logic (and introduce `getImageEtag`-function to do it) to happen inside `image-optimizer` for better consistency. ### Why? Currently when an image is requested and it becomes stale, the server will trigger a full image optimization for that image, using a significant spike in CPU-usage for routes that have multiple images (e.g. an image carousel). ### How? By calculating and storing the upstream etag in the cache entry, we can utilize the previously cached entry to check if the upstream image has remained the same, making it possible to reuse the previously cached entry again, as the image optimization would produce the same results anyways. --------- Co-authored-by: Steven <steven@ceriously.com>

Updates the test to ensure it passes with both Webpack/Turbopack.

Now you can do `x.dbg().await?` not only for `Vc<T>` but also `Value<T>`.

### Why? '[object: Object]` is not very helpful

Co-authored-by: Lee Robinson <me@leerob.io>

Co-authored-by: Lee Robinson <me@leerob.io> Co-authored-by: Delba de Oliveira <32464864+delbaoliveira@users.noreply.github.com>

### What? Known not existing exports will show an error. But only when they bindings are actually used. Unused imports won't trigger an error. ``` error - [bindings] .../invalid-export/index.js:1:0 Export Abc doesn't exist in target module + v-----------------------------v 1 + import { Abc } from "./module"; + ^-----------------------------^ 2 | // We don't report errors for unused invalid exports 3 | import { Def } from "./module"; 4 | import * as X from "./module"; 5 | The export Abc was not found in module [project]/turbopack/crates/turbopack-tests/tests/execution/turbopack/exports/invalid-export/input/invalid-export/module.js [test] (ecmascript) <exports>. Did you mean to import Abd? All exports of the module are statically known (It doesn't have dynamic exports). So it's known statically that the requested export doesn't exist. | These are the exports of the module: | Abd, Eef, a, b, c | ``` fixes PACK-3172

Empty Query Param in redirect url causes 500 error issue #68039 . Fixes: #68039 . Also added test case for same. --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

Follow up to #69850: I was unintentionally mutating the prefetch entries data property, causing the wrong params to persist on an aliased entry. This does a shallow copy on the treePatch with the updated segment data as the intent was for this to be ephemeral.

### Why? When adding `app/favicon.ico`, the `sizes` property were set to `"any"` to [avoid a browser bug](https://evilmartians.com/chronicles/how-to-favicon-in-2021-six-files-that-fit-most-needs) where an `.ico` icon is favored over `.svg`. However, we changed the behavior to set the actual size at #53343 but the doc wasn't updated together. ### How? Removed the previous explanation and updated that`sizes="any"` may be applied if the extension is `.svg` or the size is not specified. Closes #69553 --------- Co-authored-by: Jiachi Liu <inbox@huozhi.im>

### What? Allow to specify transitions via rules. ### Why? This will enable to apply a transition on every style import and allows to put style imports from RSC in the client context

## Why? We're running into the GitHub rate limit when querying for 50 `per_page`, so reducing to 25. Also adding logic throw a `info(...)` message when no issues by <xyz> version are found.

…ize (#69144) ### What? When using google fonts, we import files and write them with their hash to the filesystem. Some of these fonts get very close to the file size limit which, after appending a hash onto them, causes turbo to fail with a write error. ### Why? Different operating systems have different path length requirements. In short: - macOS: limit per segment, longer limit for entire path length - unix: limit per segment, longer limit for entire path length - windows: hard limit of 260 chars unless using UNC We need to respect these. ### How? A small change in the font code that rewrites filesystem paths if they approach this limit, with some buffer room. ~Rather than truncating I opted for a 64 bit hash because many fonts have long common prefixes. We could bump this up to 128 or higher but in practice there will be very very few files that need this code path.~ _(bgw):_ This now uses a combination of truncation (for legibility/debuggability) and a 128 bit hash (for collision avoidance). Another option is a general handler for this where reads and writes are intercepted in turbo tasks fs but I wanted to achieve a few things: - explicitly handle sources of problematic file names at the source, rather than having a blanket impl - avoid obscure hashed names appearing in our outputs generally since it makes debugging harder - avoid lots of hashing on reads and writes To me it is safer to have the source ensure they are meeting the needs of the system since we can control source specific implementations. Closes PACK-3012 --------- Co-authored-by: Benjamin Woodruff <benjamin.woodruff@vercel.com>

@samcx

This PR updates the with-xstate example to use the App Router. Here are the changes that have been made: - Renamed the `pages` folder to the `app` folder. - Updated the routing for / files to align with the App Router. - moved the `components` and `machine` folder inside the app directory by renaming it to `_components` and `_machine` - Added the `layout.tsx` file as part of the App Router. - Updated the package.json file. CC: @samcx --------- Co-authored-by: samcx <sam@vercel.com>

This removes our invalid revalidate default fallback when a revalidate value isn't returned from render as render should always return a valid value. We also already have an invariant when an invalid revalidate value is returned we just need to ensure that case could be hit properly. This also fixes some invalid test cases which were sending invalid headers. x-ref: [slack thread](https://vercel.slack.com/archives/C0676QZBWKS/p1726061828198529?thread_ts=1720714625.621179&cid=C0676QZBWKS)

### What? Improves guard detection for node.js process so it can handle code like `typeof process.off === "function" && process.off()`. ### Why? It caused several issues for sites. x-ref: https://vercel.slack.com/archives/C03S8ED1DKM/p1724702057947199 ### How?

`ResolvedVc` is a wrapper type for `Vc`. `ResolvedVc` derefs to `Vc`, but `.await`/`IntoFuture` doesn't follow the deref.

### What? Update `lightningcss` ### Why? To apply parcel-bundler/lightningcss#796 and parcel-bundler/lightningcss#797 ### How? - Closes PACK-3190 - Closes PACK-3191

### Why? Styles imported from RSC are processed twice by webpack loaders and postcss. This is not only expensive, but can also lead to hanging problems if both executions doesn't result in the same code (e. g. when using randomness in webpack loaders). ### What? Add a transition to styles to place them in the client context when import on RSC side. fixes #69643 fixes PACK-3243

### What? the template uses page.js, but the case adds a `page.tsx`, causing a duplicate page error

### What? Use strongly consistent reads to improve hmr stability

### What? In CI report unexpected errors directly instead of writing them to a file

This auto-generated PR updates font data with latest available Co-authored-by: vercel-release-bot <infra+release@vercel.com>

### What? * stable Vcs without TraitRef::cell * move entrypoints awaiting earlier * consistent middleware and instrumentation endpoint using a empty endpoint to avoid errors * handle error when subscribing to changes

### What? It should reload with the default HMR handling instead of a custom dispose method

…9944) We now use xxh3 for our deterministic non-cryptographic hashing needs. [![deadcode](https://graphite-user-uploaded-assets-prod.s3.amazonaws.com/WjlTemxb6oA4PgZFmj08/fabeefec-6b93-4740-85c6-16e9e66a8e29/deadcode.jpeg)](https://app.graphite.dev//settings/meme-library?org=vercel)

### What? remove entries from versioned content map again ### Why? We don't want to have stale entries in the map when things are removed

## Why? Issues are still getting created where folks are not testing on the latest canary. We'll add more explicit language to emphasize the need to test on canary as well. --------- Co-authored-by: Tim Neutkens <tim@timneutkens.nl>

### What? Adds a passthrough nonce for Google Analytics when inside the Google Tag Manager 3rd party script. ### Why? Promotes security by allowing users to leverage a nonce instead of `unsafe-inline` ### How? Passes the nonce through to the underlying scripts within Google Tag Manager. Closes: #61820 Fixes #61714 --------- Co-authored-by: Vincent <git@vincenttunru.com> Co-authored-by: JJ Kasper <jj@jjsweb.site>

@samcx

This PR updates the with-eslint example for using the App Router. Here are the changes that have been made: - Renamed the "pages" folder to the "app" folder. - Added the layout.tsx file as part of the App Router. - Updated the package.json file. - Updated tsconfig.json file CC: @samcx --------- Co-authored-by: samcx <sam@vercel.com>

This ensures we properly handle the `_not-found` remapping of the entry and ensures it invalidates properly when changed with the experimental trace handling. Prior to this you would see `failed to detect changes` on every successive build for the `not-found` route if present. Closes: NDX-263

Clarify that `unstable_noStore()` is only for use inside of Server Components to opt-out of Static Rendering, Client Component though initially Server Rendered cannot call this function as it is a Server-side only function Co-authored-by: JJ Kasper <jj@jjsweb.site>

Improved page copy for a better understanding of its text.  Co-authored-by: JJ Kasper <jj@jjsweb.site>

Remove stores param as new integration does not support initializing a schema. This change basically reverts back to the url that was used prior to the new integration last week, which still works just fine. Co-authored-by: JJ Kasper <jj@jjsweb.site>

Co-authored-by: JJ Kasper <jj@jjsweb.site>

…ple (#68965)  ## What? This PR fixes an incorrect usage of the `Button` component in the client component example. The example mistakenly used a `Button` component without importing or defining it properly, whereas the intent was to use the native HTML `button` element. ## Why? The original example incorrectly suggested that a custom `Button` component was being used. However, the intent was likely to demonstrate the usage of the native `button` element with an `onClick` handler to trigger a Server Action. The existing code could lead to confusion, as it implied the existence of a custom component named `Button` without any clear definition or `import`, which would also result in a recursive component reference issue. ## How? This fix ensures that the example clearly demonstrates the intended use of the native button element with an `onClick` handler, aligning with best practices and avoiding potential errors or misunderstandings in component usage. Co-authored-by: JJ Kasper <jj@jjsweb.site>

This is how I fixed the issue when this error message came up in my command line.  --------- Co-authored-by: Lee Robinson <me@leerob.io>

### Overview This PR removes the `geo` and `ip` properties from `NextRequest` for Next.js 15. These values had to be provided by your hosting provider. Since this feature was not generally useful for the broader Next.js community (outside of Vercel), we feel it makes more sense to keep this as a Vercel-specific feature and not something built into the framework. ### Alternatives If you are using Vercel, you can alternatively use the `geolocation` and `ipAddress` functions from [`@vercel/functions`](https://vercel.com/docs/functions/vercel-functions-package) instead: ```js filename="middleware.ts" import { geolocation, ipAddress } from '@vercel/functions'; import type { NextRequest } from 'next/server' export function middleware(request: NextRequest) { const { city } = geolocation(request); const ip = ipAddress(request) // ... } ``` --------- Co-authored-by: Eng Zer Jun <engzerjun@gmail.com>

OpenTelemetry and many other APM libraries rely on `import-in-the-middle` to hook ES module loading and `import-in-the-middle` does not work if it's bundled since the ESM loader hook picks up a different copy of the library to the bundled code. Nextjs uses dynamic import in some places to load modules and these can be hooked via `import-in-the-middle` but it's currently not possible to hook these modules when `import-in-the-middle` gets bundled. I'm currently working on auto instrumentation of cache handlers which are dynamically imported. You can add `import-in-the-middle` to `serverExternalPackages` but this only works if it's a direct dependency. It doesn't appear to stop bundling if `import-in-the-middle` is a sub-dependency. --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

…methods in Route Handlers (#70049) ## Why? We currently do not clarify that other supported HTTP Methods (all other besides `GET`) in Route Handlers are not cached. This remains true even if they are placed alongside a `GET` that is cached. --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

…6074) ## What? Not a case people will typically run into as you wouldn't import `.d.ts` files. However it's being imported in the React Compiler playground here with webpack customization: facebook/react#29186

Smol one that I missed!

Fix repo names so that the commands can be copy-pasted  Co-authored-by: JJ Kasper <jj@jjsweb.site>

…#70014) ### What? Fix usages of `SyntaxContext` in server actions transform. ### Why? This is part of [the tree-shaking PR](#69344). ### How?

The path that Turbopack outputs is more correct and it doesn't affect the functioning of Next.js as the path is read from the manifest.

…Map` (#69079) Building on #68698, which causes `Error`s propagated from the `VersionedContentMap` to surface as errors instead of reloading the page, allow intentionally missing values to be returned. There are scenarios in which this is desirable and not an error, such as when the content of entries change and the page must be reloaded. In these cases, a `Update:Missing` is returned, which causes the page to be reloaded. Test Plan: CI PACK-3188 --------- Co-authored-by: Tobias Koppers <tobias.koppers@googlemail.com>

) ### What? Port more linting features from `webpack/middleware-plugin.ts`. ### Why? To match the behavior of the default mode. ### How? Closes PACK-3224

…ppens (#69977) ### What? report error to stderr when an error in a subscribtion happens

This adds meta information to our diagnostics folder for easier debugging. Closes: NDX-288

…69307) In the terminal commands usage examples, the `--example` flag should be followed by the example, not by `[your-project-name]` Co-authored-by: JJ Kasper <jj@jjsweb.site>

Co-authored-by: JJ Kasper <jj@jjsweb.site>

This test case isn't guaranteed to always have the value in the output so removes for now.

Signed-off-by: haouvw <jchaodaohang@foxmail.com> Co-authored-by: JJ Kasper <jj@jjsweb.site>

#70071) …example Usage did not explain the use case that is being shown in the component example which is when fetch is not available such as using an ORM or CMS client to query your data  Co-authored-by: JJ Kasper <jj@jjsweb.site>

### Why? I've been reading the docs, and realized these little things while reading 😄 Co-authored-by: JJ Kasper <jj@jjsweb.site>

### What? Fixed an error when next#dev runs NEXT_SERVER_EVAL_SOURCE_MAPS=1 taskr in windows ### Why? Running the next.js source code library through pnpm run dev on windows causes the following error ``` next#dev > cache miss, executing 17f7b16b07333365 > next@15.0.0-canary.152 dev D:\workspace\next.js\packages\next │ > NEXT_SERVER_EVAL_SOURCE_MAPS=1 taskr │ 'NEXT_SERVER_EVAL_SOURCE_MAPS' Not an internal or external command, not a runnable program or batch file. │ ELIFECYCLE Command failed with exit code 1. │ command finished with error: command ``` ### How? Set the next.js environment variable by adding the cross-env library Closes NEXT- Fixes # --> Co-authored-by: JJ Kasper <jj@jjsweb.site>

Just adding more details here, as it was pretty bare before. --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

…already handles that (#69186) [Preview](https://github.com/vercel/next.js/blob/sebbie/08-22-_docs_recommend_disabling_static_analysis_to_combat_build_oom_iff_ci_already_handles_that/docs/02-app/01-building-your-application/06-optimizing/13-memory-usage.mdx#disable-static-analysis) Larger apps may encounter OOM in builds during static analysis (eslint+ts). Most larger apps already have a dedicated CI which is better suited for these tasks. This may result in deploy of broken builds but so is deploying without waiting for tests. The added section specifically talks about OOM and recommends manual promotion after static analysis. How to do CI/CD properly is its own industry though so we shouldn't be too opinionated in smaller sections and just describe what tools are available. Closes https://linear.app/vercel/issue/NEXT-3690

Just adding one word to this guide to add clarity; the assumption being that this is indeed true, that only *exported* functions within a file marked with `use server` are made available through API to the client. Co-authored-by: JJ Kasper <jj@jjsweb.site>

Co-authored-by: vercel-release-bot <infra+release@vercel.com>

Add "@sparticuz/chromium" to server-external-packages.json as it's now a defacto-standard for headless chromium in serverless environments.  --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

added js version of 'localized sitemap', which was missing from the md  --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

## Why? Experimental PPR does not work for Next.js v14. Technically, you can try using v14 canary, but at that point, you should just be using v15 canary because both requires React v19. - Fixes #69791 --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

--------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

This fixes a race where if the body was resolved before the clone operation, it would clone later, resulting in an error being thrown due to the body already being consumed.

This PR updates the with-web-worker example to use the App Router. Here are the changes that have been made: - Renamed the `pages` folder to the `app` folder. - Updated the routing for `/` files to align with the App Router. - Added the layout.tsx file as part of the App Router. - Updated the package.json file. The following actions were performed as part of this PR: - Ran `pnpm prettier-check` with no issues found. - Executed the `pnpm check-examples` script. --------- Co-authored-by: samcx <sam@vercel.com>

Closes #68752 --------- Co-authored-by: Sam Ko <sam@vercel.com>

Closes: #69260 Co-authored-by: Kawahara Shotaro <121674121+k-taro56@users.noreply.github.com>

…nc_trait macro) (#70084) ## What? Use rustc's relatively new native support for async traits: https://blog.rust-lang.org/2023/12/21/async-fn-rpit-in-traits.html Instead of the `async_trait` macro: https://docs.rs/async-trait/latest/async_trait/ The main differences are: - `async_trait` supports dynamic dispatch. The compiler-native version does not yet. - The compiler-native version often requires using messier `-> impl Future<...> + Send + '_` return types to declare additional bounds (usually `Send`). - `async_trait` forcibly boxes your returned `Future<...>` as a `Box<dyn Future<...>>`. ## Why? - We don't need dynamic dispatch. - Most of the `TaskInput` code is macro-generated, so the uglier `impl Future` syntax doesn't matter. - This is a hot codepath, and there's overhead associated with boxing a `dyn Future`. Removing the dynamic dispatch avoids pointer indirection and gives the compiler more opportunities to inline code. ## Benchmarks No measurable change. ``` cargo bench -p turbopack-bench -p turbopack-cli -- "hmr_to_eval/Turbopack CSR" ``` ``` bench_hmr_to_eval/Turbopack CSR/1000 modules time: [15.439 ms 15.633 ms 15.978 ms] change: [-1.3769% +1.0017% +3.3534%] (p = 0.46 > 0.05) No change in performance detected. ``` ``` TURBOPACK_BENCH_STRESS=yes cargo bench -p turbo-tasks-memory 500 ``` ``` turbo_tasks_memory_stress/fibonacci/500 time: [539.45 ms 544.10 ms 548.41 ms] thrpt: [228.39 Kelem/s 230.20 Kelem/s 232.18 Kelem/s] change: time: [-1.3718% -0.1154% +1.1538%] (p = 0.87 > 0.05) thrpt: [-1.1406% +0.1155% +1.3909%] No change in performance detected. Found 1 outliers among 20 measurements (5.00%) 1 (5.00%) low mild ```

…8648) ### What? Improve error messages about invalid CSS files. ### Why? Currently, `sass-loader` compiles CSS as a single line, and as the issue emitter prints the whole line, the error message is not good. ### How? Closes PACK-3187 Fixes #68311 --------- Co-authored-by: Tobias Koppers <tobias.koppers@googlemail.com>

This test was checking for an unsupported feature, skipped and documented.

…0016) ### What? `EcmascriptModulePartAsset` is invalid if the splitting is failed. We now carefully create it only when the original module is splittable. ### Why? This is part of [the tree-shaking PR](#69344).

### What? Improve the graph analyzer of the tree-shaking pass. This PR does **not** server-action specific changes for tree shaking. ### Why? This is part of [the tree-shaking PR](#69344). ### How?

…70103) Co-authored-by: Jiachi Liu <inbox@huozhi.im>

Closes PACK-3249 Closes #70056 Array indexing always inserts `unknown` into the result for mutable arrays (because the array might have been mutated somehow in the meantime). The same wasn't done for `Iterated` (for-of loops).

Bump sass-loader to include webpack-contrib/sass-loader#1141. Closes PACK-2447 Previously sass-loader called `this.async()()` when failing to load the dependency, which meant that the following Webpack loaders still run (because that call is passes `error = undefined` and `content = undefined`). That PR changes it to `this.async()(err)`. This in turn changes the **Turbopack** error message in this case from ``` resolve-url-loader: PostCSS error PostCSS received undefined instead of string ``` to ``` Error evaluating Node.js code Error: Cannot find module 'sass' ``` I upgraded to sass-loader 15 because 16 has a pretty breaking change regarding options: https://github.com/webpack-contrib/sass-loader/releases/tag/v16.0.0 [Breaking sass-loader changes between 12.6.0 and 15.0.0](https://github.com/webpack-contrib/sass-loader/blob/master/CHANGELOG.md): - 15.0.0 (2024-07-23) - prefer sass-embedded over sass by default - 14.0.0 (2024-01-15) - removed `fibers` support - minimum supported Node.js version is 18.12.0 - 13.0.0 - minimum supported Node.js version is 14.15.0 - emit `@warn` at-rules as webpack warnings by default, if you want to revert behavior please use the warnRuleAsWarning option

This is the only build we use in automated deploy tests and when doing ad-hoc syncs. The full matrix is still available by just re-running the workflow or manually starting one. Apple resources are scarce and scaling them up is not trivial at the moment.

The root dirs tests have been failing on Windows since #68773 so this fixes to ensure we normalize the slashes passed to `fast-glob` so it looks up as expected. x-ref: https://dev.azure.com/nextjs/next.js/_build/results?buildId=98629&view=logs&j=8af7cf9c-43a1-584d-6f5c-57bad8880974&t=7ae70e63-3625-50f4-6764-5b3e72b4bd7a

…70104) Remove TypeScript-specific non-null assertion operator (`!`) from URL parsing The non-null assertion operator (`!`) is used in TypeScript to assert that a value is not `null` or `undefined`. In plain JavaScript, this operator is not valid and can cause issues. Replaced `req.url!` with a safer approach for handling potential `null` or `undefined` values to ensure compatibility with JavaScript environments.  Co-authored-by: JJ Kasper <jj@jjsweb.site>

Updated docs, authentication section: useFormStatus and useFormState should be imported from 'react-dom' instead of 'react', or else the compiler gives the following error: ⨯ src/components/LoginForm.jsx (19:40) @ login ⨯ TypeError: (0 , react__WEBPACK_IMPORTED_MODULE_6__.useFormState) is not a function or its return value is not iterable at LoginForm (./src/components/LoginForm.jsx:23:80) at stringify (<anonymous>) at stringify (<anonymous>) digest: "258431019" 17 | 18 | export function LoginForm() { 19 | const [state, action] = useFormState(login, undefined);  Co-authored-by: JJ Kasper <jj@jjsweb.site>

When following the example for server actions for button onClick, developers will see the error below. ``` Error: Only plain objects, and a few built-ins, can be passed to Server Actions. Classes or null prototypes are not supported. ``` That is because the server action function is taking in the button onClick `event` in the action. Co-authored-by: JJ Kasper <jj@jjsweb.site>

fix: #69525 ## What I fixed an issue in the with-supabase example where duplicate className in deploy-button.tsx was causing a Tailwind CSS error. Co-authored-by: JJ Kasper <jj@jjsweb.site>

This version of Turborepo reduces the CPU usage when running `pnpm dev` (via vercel/turborepo#9121).

…70128) ### **Why?** Related desciption says "see the API reference" but does not actually link to the API reference and instead links to the exact same page.  Co-authored-by: JJ Kasper <jj@jjsweb.site>

## What? Remove `version` attribute post deprecation from the all docker compose files. ## Why? For the latest docker compose (i.e docker compose v2), the `version` attribute has been deprecated. This is currently a warning when you do docker compose up but in the future would become a requirement. [Version and name top-level elements](https://github.com/compose-spec/compose-spec/blob/main/spec.md#version-and-name-top-level-elements) mention that the `version` has become obsolete for the latest docker compose. ## How? Remove the `version` field in the compose files to stop the warning that one may get during docker builds. ### Adding or Updating Examples - [x] The "examples guidelines" are followed from our contributing doc https://github.com/vercel/next.js/blob/canary/contributing/examples/adding-examples.md - [x] Make sure the linting passes by running `pnpm build && pnpm lint`. See https://github.com/vercel/next.js/blob/canary/contributing/repository/linting.md Similar PR: #68503  --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

## Description At #60388, upgrading.md was moved. Then update the link to correct directory. ❌ [/docs/02-app/01-building-your-application/10-upgrading](https://github.com/vercel/next.js/blob/canary//docs/02-app/01-building-your-application/10-upgrading) ✔️ [/docs/02-app/01-building-your-application/11-upgrading](https://github.com/vercel/next.js/blob/canary//docs/02-app/01-building-your-application/11-upgrading) ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide x-ref: #60388  Co-authored-by: JJ Kasper <jj@jjsweb.site>

I noticed an incorrect anchor at the docs '[Route Segment Config](https://nextjs.org/docs/app/api-reference/file-conventions/route-segment-config)' so I fixed it. Co-authored-by: JJ Kasper <jj@jjsweb.site>

`export declare const sendGTMEvent: (data: Object) => void;` `sendGTMEvent` currently accept Object only. example: ``` sendGTMEvent({ event: "clicked", value: '123' }); ``` Co-authored-by: JJ Kasper <jj@jjsweb.site>

Hi there, I was trying to understand how to use [examples/github-pages](https://github.com/vercel/next.js/tree/canary/examples/github-pages) and when I followed the instructions my page was rendering without stylesheets. I found the extant README content to be clear, but after I found #38777 I was able to understand what the current content meant, so I thought it would be helpful to add this to the original example. Open to edits on content or clarity. Thanks! --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

1. `useFormState` can't be found in the current react version, for some reason. Picking it from `react-dom` works just fine. 2. Reading `state.errors.name` from the `state` object before it becomes defined returns a **TypeError**. Checking if it's defined first before calling its children works just fine.  --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

### Improving Documentation Changed example endpoint to available endpoint Co-authored-by: JJ Kasper <jj@jjsweb.site>

`charset` can be included in the `content-type` header which currently causes `RSC` detection to fail when it shouldn't so this relaxes the check to just ensure the RSC `content-type` includes the expected type not an exact match. --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

I got the error below when I tried to use the [Pending States example](https://nextjs.org/docs/app/building-your-application/data-fetching/server-actions-and-mutations#pending-states) in the documentation. After I changed the import source to react-dom, the problem was fixed. ![useFormStatus-example-warning](https://github.com/user-attachments/assets/a7ae113a-d7b4-4118-9ab7-cbd43dfd8cb1) Co-authored-by: JJ Kasper <jj@jjsweb.site>

Fixes the confusion from #69503. Co-authored-by: JJ Kasper <jj@jjsweb.site>

### What? Remove unsafe-inline from with-strict-csp example in the script-src ### Why? The main purpose of having strict CSP headers is to improve the web app security posture. On securityheaders.com the grade is capped to A if script-src contains 'unsafe-inline'. In order to reach A+. it needs to be removed. ### How? Edit the script-src in the with-strict-csp-app example Co-authored-by: JJ Kasper <jj@jjsweb.site>

- update rate limit note  Co-authored-by: JJ Kasper <jj@jjsweb.site>

### What? Minor typo on the env file name Co-authored-by: JJ Kasper <jj@jjsweb.site>

#68668)  Co-authored-by: JJ Kasper <jj@jjsweb.site>

…ocs (#68248)  ### Clarify width and height properties in Image component docs This PR updates the documentation for the `width` and `height` properties of the Image component to clarify that these attributes represent the intrinsic dimensions of the image and do not determine the rendered size, which is controlled by CSS. This clarification aligns the documentation with the behavior of the HTML `<img>` tag. Fixes issue #68228. Closes #68228 --------- Co-authored-by: JJ Kasper <jj@jjsweb.site> Co-authored-by: Willem-Jaap <67187467+Willem-Jaap@users.noreply.github.com>

### What? Fixes the error `Cannot read property 'startTime' of null` on Chrome 74 for Android 9 by ensuring the `beforeHydrationMeasure` and `hydrationMeasure` are not falsy values before accessing their properties. <img width="1164" alt="Screenshot 2024-07-02 at 11 14 53 PM" src="https://github.com/vercel/next.js/assets/52588326/34df95b6-a4fe-4886-96b8-a3c873a2ae6a"> ### Why? The application crashes on Chrome 74 for Android 9 due to `null` or other falsy values being returned in certain situations. This fix ensures that the `startTime` property is only accessed if it is not a falsy value, thereby preventing the crash. ### How? Added a falsy check before accessing the `startTime` property in the performance measurement logic. Specifically, modified the code at [this location](https://github.com/vercel/next.js/blob/b9bd23baec14508400c502b3651f4cf2497e883b/packages/next/src/client/index.tsx#L498) Co-authored-by: JJ Kasper <jj@jjsweb.site>

## Summary The error message should correspond to its value. In addition, add `?` to the state of password like the others. ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide x-ref: #70028

The TypeScript error `Props must be serializable for components in the "use client" entry file, "reset" is invalid.` for the `reset` prop in `error.tsx` and `global-error.tsx` was fixed in #46898 and #48756, respectively. However, a regression of the fix was accidentally introduced in #67211. This PR restores the previous behavior and adds fixtures for manual testing. (Since the Next.js TS plugin only applies to VSCode, manual testing in VSCode is required. See [`test/development/typescript-plugin/README.md`](https://github.com/vercel/next.js/blob/a5f30e68f8d1fe614965abc5df0234a869c6db8a/test/development/typescript-plugin/README.md) for details.)

### What? Intercepting routes should only override the slot they are in and not affect other parts of the tree. This PR makes sure we replace all other slots with the default route. Closes PACK-3219 Fixes #69299 Closes #69575

This ensures we chunk revalidate tag requests so we don't send too many at once. x-ref: [slack thread](https://vercel.slack.com/archives/C03UR7US95F/p1726585656900449?thread_ts=1726049244.204009&cid=C03UR7US95F) --------- Co-authored-by: Kelly Davis <kldavis4@users.noreply.github.com>

Co-authored-by: Jiachi Liu <inbox@huozhi.im>

This PR improves the integration test of dead code elimination of Server Reference IDs in the client bundle, to cover: - Barrel files (re-exports and nested imports) - Side effects in server boundary and intermediate modules - SWC compression

### Why? Add tests for `next-types-plugin`. The types are generated to type check the file conventions (page, layout, route). Prerequisite of #64177 Closes NDX-296 --------- Co-authored-by: Jiachi Liu <inbox@huozhi.im>

supersedes #51783 ## What This PR fixes the issue of `<a>` tags for internal routing not being caught by linting in the app directory. ## Why The linting of the html links should be consistent across pages and app directories. ## How - Added a new function to parse the URLs in the appDir (since it's a bit different from the /pages directory) - Lint now checks against pagesUrls and appUrls - Added tests for the appDir in the no-html-link-for-pages test file fixes #51742 --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

params from `{}` to `{ slug: undefined }`. Fixes #70124

…0180) Updates the docs to add a note about running `tsc` yourself. If you're ignoring Typescript errors during build time you should probably run `tsc` yourself before building. Co-authored-by: JJ Kasper <jj@jjsweb.site>

…ion (#64177) ### Why? The `next build` failed because it was trying to type check route-conventional files (page, layout, route) in the `_private` folder, which should not be considered as route segments. x-ref(CI): https://github.com/vercel/next.js/actions/runs/10907159175/job/30270204880?pr=64177#step:29:117 ### How? Don't generate type files for files that are under the `<app-dir>/**/_*` pattern. Fixes #64162 Closes NDX-283 Closes NDX-285

waiting for metadata is only necessary when `dynamicIO` is enabled. In the long run we will get rid of this waiting anyway when we have async dynamic APIs so we condition this behavior to be only when dynamicIO is on for now

…#68683)  ### What? sendGTMEvent pushes items to window.datalayer. THe GoogleTagManger (GTM) client uses this to know what events to the GTM server. currently if GTM is not ready, instead of adding events to the dataLayer, it writes a warning to the console saying GTM is not ready. The PR fixes that by defining the dataLayer if it is undefined so it can start to queue up events before GTM is ready https://developers.google.com/tag-platform/tag-manager/datalayer#how_data_layer_information_is_processed ### Why? If GTM init is slow we don't want to loose any events that we want to send before GTM init has happened we want to queue them up in the dataLayer Another use case for this behaviour, is that you can optionally initialise GoogleTagManager based on environment so you don't pollute GTM data when developing or running playright tests, but you can still evaluate what exists on the dataLayer to test if the correct events would be sent to GTM after testing interactions with the app. ### How? The PR fixes that by defining the dataLayer if it is undefined so it can start to queue up events before GTM is ready Fixes # fixes #66865 (discussion) fixes #68678 (issue) --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

`dyanmicIO` enables a semantic where any uncached IO will be excluded from prerenders. Before PPR this means any uncached IO would bail out of static generation at build time and during revalidates. Route handlers have a similar ability to be statically generated at build and revalidate time however we recently changed the behavior to only be enabled if you explicitly opt into this using an API like `export const revalidate` or `export const dynamic = "force-static"`. With dynamic IO though we don't really need to be this restrictive. If you use `unstable_cache` or `fetch(..., { cache: 'force-cache' })` you are opting into caching so when dynamicIO is on we can infer that the GET handler is valid to cache as long as there is no uncached IO. So in this commit we restore static generation when `dynamicIO` is enabled.

### What? This allows fallback pages that have Partial Prerendering enabled to be revalidated in minimal mode. ### How? When a request comes in without any `x-now-route-matches` headers and has partial prerendering enabled, it will consider it a revalidation request for the fallback page.

@samcx

This PR updates the with-videojs example to use the App Router. Here are the changes that have been made: - Renamed the pages folder to the app folder. - Updated the routing for / files to align with the App Router. - Added the layout.tsx file as part of the App Router. - Updated the package.json file. The following actions were performed as part of this PR: - Ran `pnpm prettier-check `with no issues found. - Executed the `pnpm check-examples` script. CC: @samcx --------- Co-authored-by: samcx <sam@vercel.com>

This re-adds internal env variables to allow providing page paths directly for debugging purposes. This is NOT a public API and should not be leveraged outside of internal Next.js debugging. x-ref: #67134 x-ref: [slack thread](https://vercel.slack.com/archives/C04DUD7EB1B/p1726519130567199)

### Why? Since `URL.canParse` is incompatible with older browsers like Safari 16, we added `canParseUrl` util at #70215. x-ref: https://caniuse.com/?search=URL.canParse To improve maintenance, we decided to add polyfill for `URL.canParse`. x-ref: [slack thread](https://vercel.slack.com/archives/C07BVA6HM17/p1726664071636899?thread_ts=1726620161.450109&cid=C07BVA6HM17) Closes NDX-302 --------- Co-authored-by: Jiachi Liu <inbox@huozhi.im>

Remove the dependency `graceful-fs` and types. I have looked at the project and it is only used in one file, inside this file only `fs` is imported but not the functionality provided by the library, finally it uses the native one through `graceful-fs`, I have simplified the logic and removed the unnecessary library. The retry functionality is implemented in the file and is not used from the dependency Co-authored-by: torresgol10.it <torresgol10.itd@gmail.com> Co-authored-by: JJ Kasper <jj@jjsweb.site>

Pathfinder is only needed if [the `allsorts` `outline` feature](https://docs.rs/allsorts/latest/allsorts/#cargo-features) is enabled, but it looks like we don't actually use the `outline` feature? This should avoid breakages for us whenever the nightly rust simd APIs change.

### Why? As the file paths to the build outputs of `@next/polyfill-module` & `@next/polyfill-nomodule` may not be intuitive without context, added comments that describes it.

### What? - Supersedes #69344. - Enable tree shaking for production builds ### Why? The history of #69344 is too messy ### How? Closes PACK-3215

### What? The previous approach was not completely correct and had a race condition when data was propagated by the edge was being balanced. This caused aggregated data to be incorrect. This caused some issues to stick around while they shouldn't do that. ### How? Now it locks both node at the same time to make sure to balance them in a single lockstep.

Co-authored-by: Jiachi Liu <inbox@huozhi.im>

@leerob

### What? Adds Nitrogen AIO custom image loader to docs. ### Why? Help folks find Nitrogen AIO as an option for Image @leerob --------- Co-authored-by: JJ Kasper <jj@jjsweb.site>

When calling `redirect` in a server action, the action handler will invoke a worker to fetch the RSC payload from the URL being redirected to, and will return it in the action request. This allows the redirect to be handled in a single request rather than requiring a second request to fetch the RSC payload on the URL being redirected to. As part of this special redirection handling, no action result is returned, so the action promise is resolved with `undefined`. As a result of this behavior, if your server action redirects to a page that has the same form that was just submitted, and you're leveraging `useActionState`, the updated state value will be undefined rather than being reset to the initial state. **More generally, redirect errors are handled by the client currently by simply resolving the promise with undefined, which isn't the correct behavior.** This PR will reject the server action promise with a redirect error, so that it'll be caught by our `RedirectBoundary`. Since React will remount the tree as part of the error boundary recovery, this effectively resets the form to it's initial state. If the action is rejected in a global `startTransition`, then it'll be handled by a global `error` handler, which will perform the redirect. And if it occurs outside of a transition, the redirect will be handled by a `unhandledrejection` handler. Because we're not able to apply the flight data as-is in the server action reducer (and instead defer to the navigation action), we seed the prefetch cache for the target page with the flight data from the action result so that we do not need to make another server roundtrip. We apply the same static/dynamic cache staleTime heuristics for these cache entries. Fixes #68549 Closes NDX-229

This whole test suite was disabled but only 2 tests rely on runtime logs, so this refactors the test to remove the runtime log dependency in favor of returning the output from the action and reading from the DOM.

This removes an inaccurate check that doesn't set a revalidate value if revalidate is `undefined` which can be the case for SSR app route pages. Also adds a regression test to ensure this doesn't break again. Fixes: #70213

This auto-generated PR updates the development integration test manifest used when testing Turbopack.

The `app_structure` Rust binding for Node.js is not used anywhere. It was introduced in #47737, but has never been integrated. Note: The removal is also part of #69680, but we hope to land this earlier in isolation to unblock work that builds on top of it.

### What? This fixes file path encoding differences between build and serve. Currently, file paths that get encoded via `encodeURIPath` (which uses `encodeURIComponent`): https://github.com/vercel/next.js/blob/c3475da0686eecbf9acc9b004e579a8e67288d70/packages/next/src/server/app-render/create-component-styles-and-scripts.tsx#L59-L61 But when matching files to serve, we use `encodeURI`: https://github.com/vercel/next.js/blob/c3475da0686eecbf9acc9b004e579a8e67288d70/packages/next/src/server/lib/router-utils/filesystem.ts#L191 This means that any file which contains reserved characters according to [RFC2396](https://datatracker.ietf.org/doc/html/rfc2396) will not match correctly. ``` reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" | "$" | "," ``` Closes PACK-3266

This is a pure refactoring in preparation for introducing an `app_route_loader_tree` in a future PR. The `app_page_loader_tree` collects information from the app directory for app _page_ entries, whereas the upcoming `app_route_loader_tree` will collect information from the app directory tailored to app _route_ entries.

This is a pure refactoring in preparation for introducing an `app_route_loader_tree` in a future PR. It enables reusing common logic for injecting imported modules from the app directory into both page and route entries.

This is a pure refactoring. Not every reserved file in the app directory represents a component module. Currently, `route` is the only exception, but a future PR will introduce another one.

Fixes #70216 ### What? The `redirect` function from `next/navigation` does not honor the redirect type when used in server actions. When setting `type` to `replace`, the server action pushes to history instead of replacing the route. This stems from #54458 ### Why? The server action should honor the redirect type. ### How? By setting `x-action-redirect` header to `<url>;<redirectType>`, we extract the type of redirect to be performed on the client. This is handled in a way that if the type is absent, we fallback to `push`. Demo: https://github.com/user-attachments/assets/3c84c6e3-2b08-4df3-89b2-5f9d290cc5ff Observe that once the form is submitted and the action replaced the URL with `/`, we didn't go back to the `/child` route. --------- Co-authored-by: Zack Tanner <1939140+ztanner@users.noreply.github.com>

…69992) Without these APIs, the only way to construct a `ResolvedVc` was: ``` Vc::cell(my_value).to_resolved().await? ``` However, we know when calling `Vc::cell` that the `Vc` must be resolved (since we just assigned it to a cell), so we can shortcut that and provide a better API: ``` ResolvedVc::cell(my_value) ``` For `#[turbo_tasks::value]`s, there's an equivalent `.resolved_cell` method.

docs: fix the signature of CacheHandler#revalidateTag My tests shows that the `tags` could be an array. I have traced it back to the code: https://github.com/vercel/next.js/blob/af9b229074e39617bdd32d390fdc8f6a36d7bcde/packages/next/src/server/lib/incremental-cache/index.ts#L246-L248 --------- Co-authored-by: malewis5 <matthew.a.lewis7@gmail.com> Co-authored-by: JJ Kasper <jj@jjsweb.site>

This collects the few different isThenable and isPromise definitions into one shared function

@samcx

This PR updates the with-three-js example to use the App Router. Here are the changes that have been made: - Renamed the pages folder to the app folder. - Updated the routing for `/`, `/birds` & `/boxes` files to align with the App Router. - Added the `layout.tsx` & `tsconfig.json` file as part of the App Router. - Updated the package.json file. The following actions were performed as part of this PR: - Ran `pnpm prettier-check `with no issues found. - Executed the `pnpm check-examples` script. CC: @samcx --------- Co-authored-by: samcx <sam@vercel.com>

- Fixes #49772

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[pull] canary from vercel:canary #122

[pull] canary from vercel:canary #122

Commits on Sep 5, 2024

Commits on Sep 6, 2024

Commits on Sep 7, 2024

Commits on Sep 8, 2024

Commits on Sep 9, 2024

Commits on Sep 10, 2024

Commits on Sep 11, 2024

Commits on Sep 12, 2024

Commits on Sep 13, 2024

Commits on Sep 14, 2024

Commits on Sep 16, 2024

Commits on Sep 17, 2024

Commits on Sep 18, 2024

Commits on Sep 19, 2024

Commits on Sep 20, 2024

[pull] canary from vercel:canary #122

Are you sure you want to change the base?

[pull] canary from vercel:canary #122

Commits on Sep 5, 2024

Commits on Sep 6, 2024

Commits on Sep 7, 2024

Commits on Sep 8, 2024

Commits on Sep 9, 2024

Commits on Sep 10, 2024

Commits on Sep 11, 2024

Commits on Sep 12, 2024

Commits on Sep 13, 2024

Commits on Sep 14, 2024

Commits on Sep 16, 2024

Commits on Sep 17, 2024

Commits on Sep 18, 2024

Commits on Sep 19, 2024

Commits on Sep 20, 2024