rsnext/packages/next/server/web/sandbox/context.ts

338 lines
11 KiB
TypeScript
Raw Normal View History

import type { Primitives } from 'next/dist/compiled/@edge-runtime/primitives'
Refactor Page Paths utils and Middleware Plugin (#36576) This PR brings some significant refactoring in preparation for upcoming middleware changes. Each commit can be reviewed independently, here is a summary of what each one does and the reasoning behind it: - [Move pagesDir to next-dev-server](https://github.com/javivelasco/next.js/pull/12/commits/f2fe154c007379f71c14960ddc553eaaaf786ffa) simply moves the `pagesDir` property to the dev server which is the only place where it is needed. Having it for every server is misleading. - [Move (de)normalize page path utils to a file page-path-utils.ts](https://github.com/javivelasco/next.js/pull/12/commits/27cedf087187b9632ef82a34b3af9cc4fe05d98b) Moves the functions to normalize and denormalize page paths to a single file that is intended to hold every utility function that transforms page paths. Since those are complementary it makes sense to have them together. I also added explanatory comments on why they are not idempotent and examples for input -> output that I find very useful. - [Extract removePagePathTail](https://github.com/javivelasco/next.js/pull/12/commits/6b121332aa9d3e50bd0f28b691fb7faea1b95f51) This extracts a function to remove the tail on a page path (absolute or relative). I'm sure there will be other contexts where we can use it. - [Extract getPagePaths and refactor findPageFile](https://github.com/javivelasco/next.js/pull/12/commits/cf2c7b842eebd8c02f23e79345681a794516b646) This extracts a function `getPagePaths` that is used to generate an array of paths to inspect when looking for a page file from `findPageFile`. Then it refactors such function to use it parallelizing lookups. This will allow us to print every path we look at when looking for a file which can be useful for debugging. It also adds a `flatten` helper. - [Refactor onDemandEntryHandler](https://github.com/javivelasco/next.js/pull/12/commits/4be685c37e3d1b797e929ea4f31495ed7b00e1cc) I've found this one quite difficult to understand so it is refactored to use some of the previously mentioned functions and make it easier to read. - [Extract absolutePagePath util](https://github.com/javivelasco/next.js/pull/12/commits/3bc078347426c73491a076d54ef4de977d9da073) Extracts yet another util from the `next-dev-server` that transforms an absolute path into a page name. Of course it adds comments, parameters and examples. - [Refactor MiddlewarePlugin](https://github.com/javivelasco/next.js/pull/12/commits/c595a2cc629b358cc61861a8a4848b7890d0a15b) This is the most significant change. The logic here was very hard to understand so it is totally redistributed with comments. This also removes a global variable `ssrEntries` that was deprecated in favour of module metadata added to Webpack from loaders keeping less dependencies. It also adds types and makes a clear distinction between phases where we statically analyze the code, find metadata and generate the manifest file cc @shuding @huozhi EDIT: - [Split page path utils](https://github.com/vercel/next.js/pull/36576/commits/158fb002d02887d7ce4be6747cf550a825a426eb) After seeing one of the utils was being used by the client while it was defined originally in the server, with this PR we are splitting the util into multiple files and moving it to `shared/lib` in order to make explicit that those can be also imported from client.
2022-04-30 13:19:27 +02:00
import type { WasmBinding } from '../../../build/webpack/loaders/get-module-build-info'
import {
decorateServerError,
getServerError,
} from 'next/dist/compiled/@next/react-dev-overlay/dist/middleware'
import { EDGE_UNSUPPORTED_NODE_APIS } from '../../../shared/lib/constants'
import { EdgeRuntime } from 'next/dist/compiled/edge-runtime'
import { readFileSync, promises as fs } from 'fs'
import { validateURL } from '../utils'
import { pick } from '../../../lib/pick'
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
const WEBPACK_HASH_REGEX =
/__webpack_require__\.h = function\(\) \{ return "[0-9a-f]+"; \}/g
interface ModuleContext {
runtime: EdgeRuntime<Primitives>
paths: Map<string, string>
warnedEvals: Set<string>
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
/**
* For a given path a context, this function checks if there is any module
* context that contains the path with an older content and, if that's the
* case, removes the context from the cache.
*/
export function clearModuleContext(path: string, content: Buffer | string) {
for (const [key, cache] of moduleContexts) {
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
const prev = cache?.paths.get(path)?.replace(WEBPACK_HASH_REGEX, '')
if (
typeof prev !== 'undefined' &&
prev !== content.toString().replace(WEBPACK_HASH_REGEX, '')
) {
moduleContexts.delete(key)
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
}
}
}
/**
* A Map of cached module contexts indexed by the module name. It allows
* to have a different cache scoped per module name or depending on the
* provided module key on creation.
*/
const moduleContexts = new Map<string, ModuleContext>()
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
interface ModuleContextOptions {
moduleName: string
onWarning: (warn: Error) => void
useCache: boolean
env: string[]
wasm: WasmBinding[]
}
const pendingModuleCaches = new Map<string, Promise<ModuleContext>>()
function getModuleContextShared(options: ModuleContextOptions) {
let deferredModuleContext = pendingModuleCaches.get(options.moduleName)
if (!deferredModuleContext) {
deferredModuleContext = createModuleContext(options)
pendingModuleCaches.set(options.moduleName, deferredModuleContext)
}
return deferredModuleContext
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
/**
* For a given module name this function will get a cached module
* context or create it. It will return the module context along
* with a function that allows to run some code from a given
* filepath within the context.
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
*/
export async function getModuleContext(options: ModuleContextOptions) {
let moduleContext = options.useCache
? moduleContexts.get(options.moduleName)
: await getModuleContextShared(options)
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
if (!moduleContext) {
moduleContext = await createModuleContext(options)
moduleContexts.set(options.moduleName, moduleContext)
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
}
const evaluateInContext = (filepath: string) => {
if (!moduleContext!.paths.has(filepath)) {
const content = readFileSync(filepath, 'utf-8')
try {
moduleContext?.runtime.evaluate(content)
moduleContext!.paths.set(filepath, content)
} catch (error) {
if (options.useCache) {
moduleContext?.paths.delete(options.moduleName)
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
}
throw error
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
}
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
}
return { ...moduleContext, evaluateInContext }
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
}
/**
* Create a module cache specific for the provided parameters. It includes
* a runtime context, require cache and paths cache.
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
*/
async function createModuleContext(options: ModuleContextOptions) {
const warnedEvals = new Set<string>()
const warnedWasmCodegens = new Set<string>()
const wasm = await loadWasm(options.wasm)
const runtime = new EdgeRuntime({
codeGeneration:
process.env.NODE_ENV !== 'production'
? { strings: true, wasm: false }
: undefined,
extend: (context) => {
context.process = createProcessPolyfill(options)
context.__next_eval__ = function __next_eval__(fn: Function) {
const key = fn.toString()
if (!warnedEvals.has(key)) {
const warning = getServerError(
new Error(
`Dynamic Code Evaluation (e. g. 'eval', 'new Function') not allowed in Middleware`
),
'edge-server'
)
warning.name = 'DynamicCodeEvaluationWarning'
Error.captureStackTrace(warning, __next_eval__)
warnedEvals.add(key)
options.onWarning(warning)
}
return fn()
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
context.__next_webassembly_compile__ =
function __next_webassembly_compile__(fn: Function) {
const key = fn.toString()
if (!warnedWasmCodegens.has(key)) {
const warning = getServerError(
new Error(
"Dynamic WASM code generation (e. g. 'WebAssembly.compile') not allowed in Middleware.\n" +
'Learn More: https://nextjs.org/docs/messages/middleware-dynamic-wasm-compilation'
),
'edge-server'
)
warning.name = 'DynamicWasmCodeGenerationWarning'
Error.captureStackTrace(warning, __next_webassembly_compile__)
warnedWasmCodegens.add(key)
options.onWarning(warning)
}
return fn()
}
context.__next_webassembly_instantiate__ =
async function __next_webassembly_instantiate__(fn: Function) {
const result = await fn()
// If a buffer is given, WebAssembly.instantiate returns an object
// containing both a module and an instance while it returns only an
// instance if a WASM module is given. Utilize the fact to determine
// if the WASM code generation happens.
//
// https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/WebAssembly/instantiate#primary_overload_%E2%80%94_taking_wasm_binary_code
const instantiatedFromBuffer = result.hasOwnProperty('module')
const key = fn.toString()
if (instantiatedFromBuffer && !warnedWasmCodegens.has(key)) {
const warning = getServerError(
new Error(
"Dynamic WASM code generation ('WebAssembly.instantiate' with a buffer parameter) not allowed in Middleware.\n" +
'Learn More: https://nextjs.org/docs/messages/middleware-dynamic-wasm-compilation'
),
'edge-server'
)
warning.name = 'DynamicWasmCodeGenerationWarning'
Error.captureStackTrace(warning, __next_webassembly_instantiate__)
warnedWasmCodegens.add(key)
options.onWarning(warning)
}
return result
}
const __fetch = context.fetch
context.fetch = (input: RequestInfo, init: RequestInit = {}) => {
init.headers = new Headers(init.headers ?? {})
const prevs =
init.headers.get(`x-middleware-subrequest`)?.split(':') || []
const value = prevs.concat(options.moduleName).join(':')
init.headers.set('x-middleware-subrequest', value)
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
if (!init.headers.has('user-agent')) {
init.headers.set(`user-agent`, `Next.js Middleware`)
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
if (typeof input === 'object' && 'url' in input) {
return __fetch(input.url, {
...pick(input, [
'method',
'body',
'cache',
'credentials',
'integrity',
'keepalive',
'mode',
'redirect',
'referrer',
'referrerPolicy',
'signal',
]),
...init,
headers: {
...Object.fromEntries(input.headers),
...Object.fromEntries(init.headers),
},
})
}
return __fetch(String(input), init)
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
const __Request = context.Request
context.Request = class extends __Request {
constructor(input: RequestInfo, init?: RequestInit | undefined) {
const url = typeof input === 'string' ? input : input.url
validateURL(url)
super(input, init)
}
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
const __redirect = context.Response.redirect.bind(context.Response)
context.Response.redirect = (...args) => {
validateURL(args[0])
return __redirect(...args)
}
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
for (const name of EDGE_UNSUPPORTED_NODE_APIS) {
addStub(context, name, options)
}
Object.assign(context, wasm)
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
return context
Refactor sandbox module cache (#31822) To run middleware we are using a **sandbox** that emulates the web runtime and keeps a module cache. This cache is shared for all of the modules that we run using the sandbox while there are some module-level APIs that must be scoped depending on the module we are running. One example of this is `fetch` where we want to always inject a special header that indicate the module that is performing the fetch and use it to avoid getting into infinite loops for middleware. For those cases the cached implementation will be the first one that instantiates the module and therefore we can actually get into infinite loops. This is the reason why #31800 is failing. With this PR we refactor the sandbox so that the module cache is scoped per module name. This means that one execution of a middleware will preserve its cache only for that module so that each execution will still have its own `fetch` implementation, fixing this issue. Also, with this refactor the code is more clear and we also provide an option to avoid using the cache. ## 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`
2021-11-26 13:06:41 +01:00
},
})
runtime.context.addEventListener('unhandledrejection', decorateUnhandledError)
runtime.context.addEventListener('error', decorateUnhandledError)
return {
runtime,
paths: new Map<string, string>(),
warnedEvals: new Set<string>(),
}
}
async function loadWasm(
wasm: WasmBinding[]
): Promise<Record<string, WebAssembly.Module>> {
const modules: Record<string, WebAssembly.Module> = {}
await Promise.all(
wasm.map(async (binding) => {
const module = await WebAssembly.compile(
await fs.readFile(binding.filePath)
)
modules[binding.name] = module
})
)
return modules
}
function buildEnvironmentVariablesFrom(
keys: string[]
): Record<string, string | undefined> {
const pairs = keys.map((key) => [key, process.env[key]])
const env = Object.fromEntries(pairs)
env.NEXT_RUNTIME = 'edge'
return env
}
function createProcessPolyfill(
options: Pick<ModuleContextOptions, 'env' | 'onWarning'>
) {
const env = buildEnvironmentVariablesFrom(options.env)
const processPolyfill = { env }
const overridenValue: Record<string, any> = {}
for (const key of Object.keys(process)) {
if (key === 'env') continue
Object.defineProperty(processPolyfill, key, {
get() {
emitWarning(`process.${key}`, options)
return overridenValue[key]
},
set(value) {
overridenValue[key] = value
},
enumerable: false,
})
}
return processPolyfill
}
const warnedAlready = new Set<string>()
function addStub(
context: Primitives,
name: string,
contextOptions: Pick<ModuleContextOptions, 'onWarning'>
) {
Object.defineProperty(context, name, {
get() {
emitWarning(name, contextOptions)
return undefined
},
enumerable: false,
})
}
function emitWarning(
name: string,
contextOptions: Pick<ModuleContextOptions, 'onWarning'>
) {
if (!warnedAlready.has(name)) {
const warning =
new Error(`You're using a Node.js API (${name}) which is not supported in the Edge Runtime that Middleware uses.
Learn more: https://nextjs.org/docs/api-reference/edge-runtime`)
warning.name = 'NodejsRuntimeApiInMiddlewareWarning'
contextOptions.onWarning(warning)
console.warn(warning.message)
warnedAlready.add(name)
}
}
function decorateUnhandledError(error: any) {
if (error instanceof Error) {
decorateServerError(error, 'edge-server')
}
}