Subtask of #31506
* move `serverBuildDir` and getter function of `publicDir` to next-server as only used place
* simplify `getMiddlewareInfo`, remove `distDir` and other params could be accessed from web server itself
Related to #31506, this PR adds `getPrerenderManifest` and `getRoutesManifest` methods to provide manifest files to the base server. In the future, we can consider making these methods async, or provide necessary manifests from the very top.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Part of #31506, this PR removes `loadEnvConfig` and `chalk` from the base server while keeping the same behavior for the node server.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
The image prop `onLoadingComplete()` was unexpectedly called multiple times because it uses a [callback ref](https://reactjs.org/docs/refs-and-the-dom.html#callback-refs).
This could lead to an infinite loop if `onLoadingComplete()` calls `setState()` as demonstrated in the updated test.
The solution is to handle refs with `useRef()` and `useEffect` so `onLoadingComplete()` is called at most once per component instance.
- Fixes#33463
Fixes#33462
Tried upgrading to v3, but we rely on the `timeout` property:
e5dee17f77/packages/next/telemetry/post-payload.ts (L12)
Which was removed since it's non-standard:
https://github.com/node-fetch/node-fetch/blob/main/docs/v3-UPGRADE-GUIDE.md#the-timeout-option-was-removed
I wanted to keep this PR minimal, so I did not try to work around the above.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
We currently have inconsistencies when working with URLs in the Edge Functions runtime, this PR addresses them introducing a warning for inconsistent usage that will break in the future. Here is the reasoning.
### The Browser
When we are in a browser environment there is a fixed location stored at `globalThis.location`. Then, if one tries to build a request with a relative URL it will work using that location global hostname as _base_ to construct its URL. For example:
```typescript
// https://nextjs.org
new Request('/test').url; // https://nextjs.org/test
Response.redirect('/test').headers.get('Location'); // https://nextjs.org/test
```
However, if we attempt to run the same code from `about:blank` it would not work because the global to use as a base `String(globalThis.location)` is not a valid URL. Therefore a call to `Response.redirect('/test')` or `new Response('/test')` would fail.
### Edge Functions Runtime
In Next.js Edge Functions runtime the situation is slightly different from a browser. Say that we have a root middleware (`pages/_middleware`) that gets invoked for every page. In the middleware file we expose the handler function and also define a global variable that we mutate on every request:
```typescript
// pages/_middleware
let count = 0;
export function middleware(req: NextRequest) {
console.log(req.url);
count += 1;
}
```
Currently we cache the module scope in the runtime so subsequent invocations would hold the same globals and the module would not be evaluated again. This would make the counter to increment for each request that the middleware handles. It is for this reason that we **can't have a global location** that changes across different invocations. Each invocation of the same function uses the same global which also holds primitives like `URL` or `Request` so changing an hypothetical `globalThis.location` per request would affect concurrent requests being handled.
Then, it is not possible to use relative URLs in the same way the browser does because we don't have a global to rely on to use its host to compose a URL from a relative path.
### Why it works today
We are **not** validating what is provided to, for example, `NextResponse.rewrite()` nor `NextResponse.redirect()`. We simply create a `Response` instance that adds the corresponding header for the rewrite or the redirect. Then it is **the consumer** the one that composes the final destination based on the request. Theoretically you can pass any value and it would fail on redirect but won't validate the input.
Of course this is inconsistent because it doesn't make sense that `NextResponse.rewrite('/test')` works but `fetch(new NextRequest('/test'))` does not. Also we should validate what is provided. Finally, we want to be consistent with the way a browser behaves so `new Request('/test')` _should_ not work if there is no global location which we lack.
### What this PR does
We will have to deprecate the usage of relative URLs in the previously mentioned scenarios. In preparation for it, this PR adds a validation function in those places where it will break in the future, printing a warning with a link that points to a Next.js page with an explanation of the issue and ways to fix it.
Although middleware changes are not covered by semver, we will roll this for some time to make people aware that this change is coming. Then after a reasonable period of time we can remove the warning and make the code fail when using relative URLs in the previously exposed scenarios.
Part of #31506, this PR moves the code of middleware handling from the base server to the node server.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
## Bug
Fixes: #32515
Previously, we render the `suffix` after consuming 1st chunk, instead we should render it after stream finished
- [x] Related issues linked using `fixes #number`
- [x] Integration tests added
- [x] Errors have helpful link attached, see `contributing.md`
This fixes our check for fallback pages during `next export` as we are currently erroring even when the erroneous pages are not included in the `exportPathMap`. This also adds additional tests to certify the expected behavior for the error.
## Bug
- [x] Related issues linked using `fixes #number`
- [x] Integration tests added
- [x] Errors have helpful link attached, see `contributing.md`
Fixes: https://github.com/vercel/next.js/issues/29135
Adds base http classes, along with Node + Web (partial) implementations
Removes usage of IncomingMessage and ServerResponse from base server
Co-authored-by: Shu Ding <3676859+shuding@users.noreply.github.com>
* Update yarn PnP tests and disable swc file reading for PnP
* update job
* Update test
* add env variable
* update destory
* test one
* bump timeout
* update pnp install command
* only run pnp test
* add more logs
* handle exit signal
* dont inherit stdio for install
* update server start
* re-add test type
* add build log
* additional logging
* update build command
* remove separate timeout
* update install command
* install separate for better time info
* add cache pre-warming
* update yarn config
* enable other pnp tests
* Separate out tests
* fix-lint
* update path
* update test concurrency for isolated tests
* update retries
* Revert "update test concurrency for isolated tests"
This reverts commit 3a6e924df8ec61d55d3ee8a58d24cd50f0141195.
* re-enable production tests
* apply suggestions
## Feature
Resolves#30996
- [x] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [x] Related issues linked using `fixes #number`
- [x] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
include a few lines of stack trace in the full refresh overlay
error is not necessary webpack related
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
This PR is a bit big because AST definitions are modified recently because previous AST defs were too error-prone.
It will prevent plugin authors from making some common mistakes.
- Closes#33088
- Closes#31084
- Closes#33283
When there is a DOM element with id of `process`, the DOM marks it as a global, so `window.process` would exist. We should check for `process.env` to make sure it is available too.
Fixes#33135
I also tried reducing the number of variants.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
and text is easier to gzip
and it avoid the reference to `Buffer` from next
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
This PR moves `require` and `fs` usage from the base server to the node server. Closes#32322.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Production middlewares will only expose env vars that are statically analyzable, as mentioned here: https://nextjs.org/docs/api-reference/next/server#how-do-i-access-environment-variables
This creates some incompatibility with `next dev` and `next start`, where all `process.env` data is shared and can lead to unexpected behavior in runtime.
This PR fixes it by limiting the data in `process.env` with the inferred env vars from the code usage. I believe the test speaks for itself 🕺
<!--
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
-->
* JSON.stringify generic errors
* Add util for normalizing errors
* lint-fix
* Add better error for null case as well
Co-authored-by: Michael Ozeryansky <mozeryansky@users.noreply.github.com>
x-ref [#31552 ](https://github.com/vercel/next.js/issues/31552)
x-ref https://github.com/vercel/next.js/issues/32115
x-ref https://github.com/vercel/next.js/issues/32546
x-ref https://github.com/vercel/next.js/issues/32721
Since this PR https://github.com/vercel/next.js/pull/31455 is merged, `enhanced-resolve` dependency's resolved field is changed which caused broken yarn pnp.
I am not sure how this field has been changed or this is intentional or not
When I install webpack locally, `enhanced-resolve`'s resolved field in lock file is always `registry.yarnpkg.com` not `codeload.github.com`
## Bug
- [x] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Co-authored-by: Tobias Koppers <1365881+sokra@users.noreply.github.com>
Co-authored-by: JJ Kasper <22380829+ijjk@users.noreply.github.com>
* next-swc: fix ssg code elimination when used in render
* simplify logic and add more thorough tests
* comment
* remove fold_export_default and fix JSX member expression
After discussing with @sokra, seems that the proposed solution is split in two:
* We need to make sure that the `process` polyfill uses `global.process` if available. This is because middlewares are bundled using `browser` target and therefore `process.env.MY_ENV` gets shimmed into `require('process').env.MY_ENV`.
* Allow `process.env` to be statically analyzed for dependencies so they will be exported to the manifest.
Related issues:
* should fix#33043.
it should not leak into the client runtime
cc @javivelasco
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Something [between `11.0.2-canary.5` and `11.0.2-canary.6`](https://github.com/vercel/next.js/compare/v11.0.2-canary.5...v11.0.2-canary.6) changed the behavior that logged any runtime errors in `getStaticProps` to stderr. This is only observable if `getStaticProps` has a `revalidate` value, and the build did not fail. The error has to happen in a subsequent revalidation step.
This PR reverts the change and fixes#30375.
## Bug
- [x] Related issues linked using `fixes #number`
- [x] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Co-authored-by: JJ Kasper <22380829+ijjk@users.noreply.github.com>
This PR adds a new command to the `next` CLI.
Running `next info` will print useful information to the terminal about how/where Next.js is run. This information can be added to the Bug report when opening an issue in the repository.
This makes reporting issues more accurate and doesn't require the user to guess certain details, the command will retrieve it on their behalf.
Example output:
```sh
$ npx --no-install next info
Operating System:
Platform: linux
Version: #22-Ubuntu SMP Fri Nov 5 13:21:36 UTC 2021
Binaries:
Node: 16.13.0
npm: 8.1.0
Yarn: 1.22.17
pnpm: 6.24.2
Relevant packages:
next: 12.0.8-canary.14
react: 17.0.2
react-dom: 17.0.2
```
The idea is based on #32858
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [x] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Co-authored-by: Steven <229881+styfle@users.noreply.github.com>
Fixes https://github.com/vercel/next.js/issues/32024
Fixing makeStylesheetInert.
Fixes log import.
## Bug
- [x] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Co-authored-by: Devin Wall <87490815+DevinWallKcl@users.noreply.github.com>
Fixes#32727
With this PR we introduce a new header that can be used to respond from Middleware `x-middleware-cache`. When the value of this header is set to `no-cache`, the client will **not** store the effects read from a preflight response to be used in an upcoming check.
Instead of using `Cache-Control` we are using a custom header to not mess with browser specific caching. Accepting a specific value to opt out (`no-cache`) opens the future opportunity of having other caching strategies.
This feature solves the issue of having a preflight request whose parameters can change from the client affecting the effects. For example, having a cookie that would make a middleware rewrite to one pathname or another and allowing to change that cookie from the client. In that case we'd always need to revalidate the effects on navigation synchronously.
Of course when using this feature it is possible that we add some latency on navigation so the preferred mechanism will be caching by default since it covers the most common use cases.
When the URL already contains a hash (like `/#section`) and a `Link` with `href="/"` is clicked, the page should scroll to the top on the first click. Currently, it only happens with the second click (after `#section` has been removed from the URL).
Fixes#32931
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
* Move resolve-url-loader into Next.js
Fixes#32157
Moves resolve-url-loader into Next.js and strips out all features that are not used like `rework` support. Will reduce install size as well as allow for optimizing the approach in the near future.
* Update precompiled
* Use loader-utils 2
* Update trace test
* Revert "Update trace test"
This reverts commit 7c09a07871cc0ab72d5fcd4151a2d8efbc1aad8f.
* Add es5-ext as it's used in trace tests
* Update join-function.js
* Update bundle5.js
Uses a simple counter for span ids, previously these required to be randomly generated but I've changed the importer script to ensure it gets prepended `0`s to make sure it gets to 16 characters which Jaeger requires.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Noscript is not required for Image that are loaded immediately
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [x] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
Co-authored-by: Steven <229881+styfle@users.noreply.github.com>
This fixes broken links in the eslint output by removing the trailing full stop.
It also makes the formatting of (the output of) the various rules consistent.
## Documentation / Examples
- [x] Make sure the linting passes by running `yarn lint`
> I don't think this is a bug, nor a feature, nor is it really documentation.
> It's just a small nuisance that I bumped into and felt compelled to fix.
> I went with documentation as that seems the closest match
## What does this pull request do?
The elslint output of `eslint-plugin-next` contains useful links to the documentation about the various rules.
Unfortunately, on most (but not all) rules, those links are immediately followed by a full stop (`.`).
The terminal (or any parser) has no way of knowing that the full stop is not part of the URL.
So it includes it and clicking the link leads to a 404 on the nextjs.org website.
![eslint](https://user-images.githubusercontent.com/1708494/147452577-43ad4ce7-df75-4d48-ab78-70b9b8212b7e.png)
This PR fixes that by removing the full stop.
## But a final full stop is better grammar
I considered alternatives (such as [a zero-width space character](https://en.wikipedia.org/wiki/Zero-width_space#Prohibited_in_URLs)) in case the final full stop was part of the style guide or something.
However, as I went through the eslint rules, I notices that the messages for various rules were formatted inconsistently.
Some with final full stop, some without.
As such, I made the all consistent with this structure:
> [message]. See: [url]
I feel this is a better solution than using the zero-width space as these sort of invisible characters
in code can be a red flag that something fishy is going on.
I submit this pull request in the hope it will be useful, and a positive contribution to a project I have a great deal of appreciation for.
That being said, I fully understand if people would consider this a non-issue.
When an `a` tag is used to link to an internal page, but the `target="_blank"` attribute is present, ESLint should not report it as an error and should not enforce using `next/link`.
Fixes#28547
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
We didn't port this plugin to next-swc given that the eslint setup warns for this automatically. Given that we still get issues opened for this particular warning coming from the plugin it'd be good to remove it.
## Bug
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Errors have helpful link attached, see `contributing.md`
## Feature
- [ ] Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR.
- [ ] Related issues linked using `fixes #number`
- [ ] Integration tests added
- [ ] Documentation added
- [ ] Telemetry added. In case of a feature if it's used or not.
- [ ] Errors have helpful link attached, see `contributing.md`
## Documentation / Examples
- [ ] Make sure the linting passes by running `yarn lint`
We introduced the `StyleRegistry` on server side to make it concurrent safe for requests. For client we can have only 1 registry since there's only 1 instance per browser window. Keep consistence with previous usage
Closes vercel/next.js#32417
Closes vercel/next.js#30377
Revert change in #31357, it's not required anymore
* Don't render `next/head` in _error for RSC
* Save an error in `__next_s` for client to be aware RSC rendering is errored
* Fix `req.url` when errored
* Refactor some changes in middleware ssr loader
It adds AbortController and AbortSignal Web runtimes APIs to be used by the user at Edge Functions.
For doing that it delegates into `abort-controller` dependency that has been frozen to prevent any modification.
Co-authored-by: Zhang Zhi <20026577+fytriht@users.noreply.github.com>
This PR is a follow up to PR #32623 to fix the remaining blur styles.
These are unlikely to be overridden by the user but this PR is necessary to make `placeholder=empty` (default behavior) and `placeholder=blur` end up with the same inline styles on the loaded image.
Related to https://github.com/vercel/next.js/issues/18398#issuecomment-997114428