Node.js v14.14.0 文档


ECMAScript 模块#

稳定性: 1 - 实验

介绍#

ECMAScript modules are the official standard format to package JavaScript code for reuse. Modules are defined using a variety of import and export statements.

The following example of an ES module exports a function:

// addTwo.mjs
function addTwo(num) {
  return num + 2;
}

export { addTwo };

The following example of an ES module imports the function from addTwo.mjs:

// app.mjs
import { addTwo } from './addTwo.mjs';

// Prints: 6
console.log(addTwo(4));

Node.js fully supports ECMAScript modules as they are currently specified and provides limited interoperability between them and the existing module format, CommonJS.

Node.js contains support for ES Modules based upon the Node.js EP for ES Modules and the ECMAScript-modules implementation.

Expect major changes in the implementation including interoperability support, specifier resolution, and default behavior.

启用#

Node.js treats JavaScript code as CommonJS modules by default. Authors can tell Node.js to treat JavaScript code as ECMAScript modules via the .mjs file extension, the package.json "type" field, or the --input-type flag. See Modules: Packages for more details.

#

This section was moved to Modules: Packages.

import 说明符#

术语#

The specifier of an import statement is the string after the from keyword, e.g. 'path' in import { sep } from 'path'. Specifiers are also used in export from statements, and as the argument to an import() expression.

There are four types of specifiers:

  • Bare specifiers like 'some-package'. They refer to an entry point of a package by the package name.

  • Deep import specifiers like 'some-package/lib/shuffle.mjs'. They refer to a path within a package prefixed by the package name.

  • Relative specifiers like './startup.js' or '../config.mjs'. They refer to a path relative to the location of the importing file.

  • Absolute specifiers like 'file:///opt/nodejs/config.js'. They refer directly and explicitly to a full path.

Bare specifiers, and the bare specifier portion of deep import specifiers, are strings; but everything else in a specifier is a URL.

file:, node:, and data: URLs are supported. A specifier like 'https://example.com/app.js' may be supported by browsers but it is not supported in Node.js.

Specifiers may not begin with / or //. These are reserved for potential future use. The root of the current volume may be referenced via file:///.

node: 导入#

node: URLs are supported as a means to load Node.js builtin modules. This URL scheme allows for builtin modules to be referenced by valid absolute URL strings.

import fs from 'node:fs/promises';

data: 导入#

data: URLs are supported for importing with the following MIME types:

  • text/javascript for ES Modules
  • application/json for JSON
  • application/wasm for Wasm

data: URLs only resolve Bare specifiers for builtin modules and Absolute specifiers. Resolving Relative specifiers does not work because data: is not a special scheme. For example, attempting to load ./foo from data:text/javascript,import "./foo"; fails to resolve because there is no concept of relative resolution for data: URLs. An example of a data: URLs being used is:

import 'data:text/javascript,console.log("hello!");';
import _ from 'data:application/json,"world!"';

import.meta#

The import.meta metaproperty is an Object that contains the following property:

  • url <string> The absolute file: URL of the module.

ES 模块和 CommonJS 之间的差异#

强制的文件扩展名#

A file extension must be provided when using the import keyword. Directory indexes (e.g. './startup/index.js') must also be fully specified.

This behavior matches how import behaves in browser environments, assuming a typically configured server.

没有 NODE_PATH#

NODE_PATH is not part of resolving import specifiers. Please use symlinks if this behavior is desired.

没有 require、exports、module.exports、__filename、__dirname#

These CommonJS variables are not available in ES modules.

require can be imported into an ES module using module.createRequire().

Equivalents of __filename and __dirname can be created inside of each file via import.meta.url.

import { fileURLToPath } from 'url';
import { dirname } from 'path';

const __filename = fileURLToPath(import.meta.url);
const __dirname = dirname(__filename);

没有 require.resolve#

Former use cases relying on require.resolve to determine the resolved path of a module can be supported via import.meta.resolve, which is experimental and supported via the --experimental-import-meta-resolve flag:

(async () => {
  const dependencyAsset = await import.meta.resolve('component-lib/asset.css');
})();

import.meta.resolve also accepts a second argument which is the parent module from which to resolve from:

(async () => {
  // Equivalent to import.meta.resolve('./dep')
  await import.meta.resolve('./dep', import.meta.url);
})();

This function is asynchronous because the ES module resolver in Node.js is asynchronous. With the introduction of Top-Level Await, these use cases will be easier as they won't require an async function wrapper.

没有 `require.extensions`#

require.extensions is not used by import. The expectation is that loader hooks can provide this workflow in the future.

没有 `require.cache`#

require.cache is not used by import. It has a separate cache.

基于 URL 的路径#

ES modules are resolved and cached based upon URL semantics. This means that files containing special characters such as # and ? need to be escaped.

Modules are loaded multiple times if the import specifier used to resolve them has a different query or fragment.

import './foo.mjs?query=1'; // loads ./foo.mjs with query of "?query=1"
import './foo.mjs?query=2'; // loads ./foo.mjs with query of "?query=2"

For now, only modules using the file: protocol can be loaded.

与 CommonJS 的互操作性#

require#

require always treats the files it references as CommonJS. This applies whether require is used the traditional way within a CommonJS environment, or in an ES module environment using module.createRequire().

To include an ES module into CommonJS, use import().

import 声明#

An import statement can reference an ES module or a CommonJS module. import statements are permitted only in ES modules. For similar functionality in CommonJS, see import().

When importing CommonJS modules, the module.exports object is provided as the default export. Named exports may be available, provided by static analysis as a convenience for better ecosystem compatibility.

Additional experimental flags are available for importing Wasm modules or JSON modules. For importing native modules or JSON modules unflagged, see module.createRequire().

The specifier of an import statement (the string after the from keyword) can either be an URL-style relative path like './file.mjs' or a package name like 'fs'.

Like in CommonJS, files within packages can be accessed by appending a path to the package name; unless the package’s package.json contains an "exports" field, in which case files within packages need to be accessed via the path defined in "exports".

import { sin, cos } from 'geometry/trigonometry-functions.mjs';

import() 表达式#

Dynamic import() is supported in both CommonJS and ES modules. It can be used to include ES module files from CommonJS code.

CommonJS 命名空间#

CommonJS modules consist of a module.exports object which can be of any type.

When importing a CommonJS module, it can be reliably imported using the ES module default import or its corresponding sugar syntax:

import { default as cjs } from 'cjs';

// The following import statement is "syntax sugar" (equivalent but sweeter)
// for `{ default as cjsSugar }` in the above import statement:
import cjsSugar from 'cjs';

console.log(cjs);
console.log(cjs === cjsSugar);
// Prints:
//   <module.exports>
//   true

The ECMAScript Module Namespace representation of a CommonJS module is always a namespace with a default export key pointing to the CommonJS module.exports value.

This Module Namespace Exotic Object can be directly observed either when using import * as m from 'cjs' or a dynamic import:

import * as m from 'cjs';
console.log(m);
console.log(m === await import('cjs'));
// Prints:
//   [Module] { default: <module.exports> }
//   true

For better compatibility with existing usage in the JS ecosystem, Node.js in addition attempts to determine the CommonJS named exports of every imported CommonJS module to provide them as separate ES module exports using a static analysis process.

For example, consider a CommonJS module written:

// cjs.cjs
exports.name = 'exported';

The preceding module supports named imports in ES modules:

import { name } from './cjs.cjs';
console.log(name);
// Prints: 'exported'

import cjs from './cjs.cjs';
console.log(cjs);
// Prints: { name: 'exported' }

import * as m from './cjs.cjs';
console.log(m);
// Prints: [Module] { default: { name: 'exported' }, name: 'exported' }

As can be seen from the last example of the Module Namespace Exotic Object being logged, the name export is copied off of the module.exports object and set directly on the ES module namespace when the module is imported.

Live binding updates or new exports added to module.exports are not detected for these named exports.

The detection of named exports is based on common syntax patterns but does not always correctly detect named exports. In these cases, using the default import form described above can be a better option.

Named exports detection covers many common export patterns, reexport patterns and build tool and transpiler outputs. See cjs-module-lexer for the exact semantics implemented.

内置模块#

Core modules provide named exports of their public API. A default export is also provided which is the value of the CommonJS exports. The default export can be used for, among other things, modifying the named exports. Named exports of builtin modules are updated only by calling module.syncBuiltinESMExports().

import EventEmitter from 'events';
const e = new EventEmitter();
import { readFile } from 'fs';
readFile('./foo.txt', (err, source) => {
  if (err) {
    console.error(err);
  } else {
    console.log(source);
  }
});
import fs, { readFileSync } from 'fs';
import { syncBuiltinESMExports } from 'module';

fs.readFileSync = () => Buffer.from('Hello, ESM');
syncBuiltinESMExports();

fs.readFileSync === readFileSync;

CommonJS、JSON 和原生模块#

CommonJS, JSON, and native modules can be used with module.createRequire().

// cjs.cjs
module.exports = 'cjs';

// esm.mjs
import { createRequire } from 'module';

const require = createRequire(import.meta.url);

const cjs = require('./cjs.cjs');
cjs === 'cjs'; // true

实验的 JSON 模块#

Currently importing JSON modules are only supported in the commonjs mode and are loaded using the CJS loader. WHATWG JSON modules specification are still being standardized, and are experimentally supported by including the additional flag --experimental-json-modules when running Node.js.

When the --experimental-json-modules flag is included, both the commonjs and module mode use the new experimental JSON loader. The imported JSON only exposes a default. There is no support for named exports. A cache entry is created in the CommonJS cache to avoid duplication. The same object is returned in CommonJS if the JSON module has already been imported from the same path.

Assuming an index.mjs with

import packageConfig from './package.json';

The --experimental-json-modules flag is needed for the module to work.

node index.mjs # fails
node --experimental-json-modules index.mjs # works

实验的 Wasm 模块#

Importing Web Assembly modules is supported under the --experimental-wasm-modules flag, allowing any .wasm files to be imported as normal modules while also supporting their module imports.

This integration is in line with the ES Module Integration Proposal for Web Assembly.

For example, an index.mjs containing:

import * as M from './module.wasm';
console.log(M);

executed under:

node --experimental-wasm-modules index.mjs

would provide the exports interface for the instantiation of module.wasm.

实验的顶层 await#

The await keyword may be used in the top level (outside of async functions) within modules as per the ECMAScript Top-Level await proposal.

Assuming an a.mjs with

export const five = await Promise.resolve(5);

And a b.mjs with

import { five } from './a.mjs';

console.log(five); // Logs `5`
node b.mjs # works

实验的加载器#

Note: This API is currently being redesigned and will still change.

To customize the default module resolution, loader hooks can optionally be provided via a --experimental-loader ./loader-name.mjs argument to Node.js.

When hooks are used they only apply to ES module loading and not to any CommonJS modules loaded.

钩子#

resolve(specifier, context, defaultResolve)#

Note: The loaders API is being redesigned. This hook may disappear or its signature may change. Do not rely on the API described below.

The resolve hook returns the resolved file URL for a given module specifier and parent URL. The module specifier is the string in an import statement or import() expression, and the parent URL is the URL of the module that imported this one, or undefined if this is the main entry point for the application.

The conditions property on the context is an array of conditions for Conditional exports that apply to this resolution request. They can be used for looking up conditional mappings elsewhere or to modify the list when calling the default resolution logic.

The current package exports conditions are always in the context.conditions array passed into the hook. To guarantee default Node.js module specifier resolution behavior when calling defaultResolve, the context.conditions array passed to it must include all elements of the context.conditions array originally passed into the resolve hook.

/**
 * @param {string} specifier
 * @param {{
 *   conditions: !Array<string>,
 *   parentURL: !(string | undefined),
 * }} context
 * @param {Function} defaultResolve
 * @returns {Promise<{ url: string }>}
 */
export async function resolve(specifier, context, defaultResolve) {
  const { parentURL = null } = context;
  if (Math.random() > 0.5) { // Some condition.
    // For some or all specifiers, do some custom logic for resolving.
    // Always return an object of the form {url: <string>}.
    return {
      url: parentURL ?
        new URL(specifier, parentURL).href :
        new URL(specifier).href,
    };
  }
  if (Math.random() < 0.5) { // Another condition.
    // When calling `defaultResolve`, the arguments can be modified. In this
    // case it's adding another value for matching conditional exports.
    return defaultResolve(specifier, {
      ...context,
      conditions: [...context.conditions, 'another-condition'],
    });
  }
  // Defer to Node.js for all other specifiers.
  return defaultResolve(specifier, context, defaultResolve);
}

getFormat(url, context, defaultGetFormat)#

Note: The loaders API is being redesigned. This hook may disappear or its signature may change. Do not rely on the API described below.

The getFormat hook provides a way to define a custom method of determining how a URL should be interpreted. The format returned also affects what the acceptable forms of source values are for a module when parsing. This can be one of the following:

formatDescriptionAcceptable Types For source Returned by getSource or transformSource
'builtin'Load a Node.js builtin moduleNot applicable
'commonjs'Load a Node.js CommonJS moduleNot applicable
'json'Load a JSON file{ string, ArrayBuffer, TypedArray }
'module'Load an ES module{ string, ArrayBuffer, TypedArray }
'wasm'Load a WebAssembly module{ ArrayBuffer, TypedArray }

Note: These types all correspond to classes defined in ECMAScript.

Note: If the source value of a text-based format (i.e., 'json', 'module') is not a string, it is converted to a string using util.TextDecoder.

/**
 * @param {string} url
 * @param {Object} context (currently empty)
 * @param {Function} defaultGetFormat
 * @returns {Promise<{ format: string }>}
 */
export async function getFormat(url, context, defaultGetFormat) {
  if (Math.random() > 0.5) { // Some condition.
    // For some or all URLs, do some custom logic for determining format.
    // Always return an object of the form {format: <string>}, where the
    // format is one of the strings in the preceding table.
    return {
      format: 'module',
    };
  }
  // Defer to Node.js for all other URLs.
  return defaultGetFormat(url, context, defaultGetFormat);
}

getSource(url, context, defaultGetSource)#

Note: The loaders API is being redesigned. This hook may disappear or its signature may change. Do not rely on the API described below.

The getSource hook provides a way to define a custom method for retrieving the source code of an ES module specifier. This would allow a loader to potentially avoid reading files from disk.

/**
 * @param {string} url
 * @param {{ format: string }} context
 * @param {Function} defaultGetSource
 * @returns {Promise<{ source: !(string | SharedArrayBuffer | Uint8Array) }>}
 */
export async function getSource(url, context, defaultGetSource) {
  const { format } = context;
  if (Math.random() > 0.5) { // Some condition.
    // For some or all URLs, do some custom logic for retrieving the source.
    // Always return an object of the form {source: <string|buffer>}.
    return {
      source: '...',
    };
  }
  // Defer to Node.js for all other URLs.
  return defaultGetSource(url, context, defaultGetSource);
}

transformSource(source, context, defaultTransformSource)#

Note: The loaders API is being redesigned. This hook may disappear or its signature may change. Do not rely on the API described below.

The transformSource hook provides a way to modify the source code of a loaded ES module file after the source string has been loaded but before Node.js has done anything with it.

If this hook is used to convert unknown-to-Node.js file types into executable JavaScript, a resolve hook is also necessary in order to register any unknown-to-Node.js file extensions. See the transpiler loader example below.

/**
 * @param {!(string | SharedArrayBuffer | Uint8Array)} source
 * @param {{
 *   format: string,
 *   url: string,
 * }} context
 * @param {Function} defaultTransformSource
 * @returns {Promise<{ source: !(string | SharedArrayBuffer | Uint8Array) }>}
 */
export async function transformSource(source, context, defaultTransformSource) {
  const { url, format } = context;
  if (Math.random() > 0.5) { // Some condition.
    // For some or all URLs, do some custom logic for modifying the source.
    // Always return an object of the form {source: <string|buffer>}.
    return {
      source: '...',
    };
  }
  // Defer to Node.js for all other sources.
  return defaultTransformSource(source, context, defaultTransformSource);
}

getGlobalPreloadCode()#

Note: The loaders API is being redesigned. This hook may disappear or its signature may change. Do not rely on the API described below.

Sometimes it might be necessary to run some code inside of the same global scope that the application runs in. This hook allows the return of a string that is run as sloppy-mode script on startup.

Similar to how CommonJS wrappers work, the code runs in an implicit function scope. The only argument is a require-like function that can be used to load builtins like "fs": getBuiltin(request: string).

If the code needs more advanced require features, it has to construct its own require using module.createRequire().

/**
 * @returns {string} Code to run before application startup
 */
export function getGlobalPreloadCode() {
  return `\
globalThis.someInjectedProperty = 42;
console.log('I just set some globals!');

const { createRequire } = getBuiltin('module');

const require = createRequire(process.cwd() + '/<preload>');
// [...]
`;
}

示例#

The various loader hooks can be used together to accomplish wide-ranging customizations of Node.js’ code loading and evaluation behaviors.

HTTPS 加载器#

In current Node.js, specifiers starting with https:// are unsupported. The loader below registers hooks to enable rudimentary support for such specifiers. While this may seem like a significant improvement to Node.js core functionality, there are substantial downsides to actually using this loader: performance is much slower than loading files from disk, there is no caching, and there is no security.

// https-loader.mjs
import { get } from 'https';

export function resolve(specifier, context, defaultResolve) {
  const { parentURL = null } = context;

  // Normally Node.js would error on specifiers starting with 'https://', so
  // this hook intercepts them and converts them into absolute URLs to be
  // passed along to the later hooks below.
  if (specifier.startsWith('https://')) {
    return {
      url: specifier
    };
  } else if (parentURL && parentURL.startsWith('https://')) {
    return {
      url: new URL(specifier, parentURL).href
    };
  }

  // Let Node.js handle all other specifiers.
  return defaultResolve(specifier, context, defaultResolve);
}

export function getFormat(url, context, defaultGetFormat) {
  // This loader assumes all network-provided JavaScript is ES module code.
  if (url.startsWith('https://')) {
    return {
      format: 'module'
    };
  }

  // Let Node.js handle all other URLs.
  return defaultGetFormat(url, context, defaultGetFormat);
}

export function getSource(url, context, defaultGetSource) {
  // For JavaScript to be loaded over the network, we need to fetch and
  // return it.
  if (url.startsWith('https://')) {
    return new Promise((resolve, reject) => {
      get(url, (res) => {
        let data = '';
        res.on('data', (chunk) => data += chunk);
        res.on('end', () => resolve({ source: data }));
      }).on('error', (err) => reject(err));
    });
  }

  // Let Node.js handle all other URLs.
  return defaultGetSource(url, context, defaultGetSource);
}
// main.mjs
import { VERSION } from 'https://coffeescript.org/browser-compiler-modern/coffeescript.js';

console.log(VERSION);

With the preceding loader, running node --experimental-loader ./https-loader.mjs ./main.mjs prints the current version of CoffeeScript per the module at the URL in main.mjs.

转译器加载器#

Sources that are in formats Node.js doesn’t understand can be converted into JavaScript using the transformSource hook. Before that hook gets called, however, other hooks need to tell Node.js not to throw an error on unknown file types; and to tell Node.js how to load this new file type.

This is less performant than transpiling source files before running Node.js; a transpiler loader should only be used for development and testing purposes.

// coffeescript-loader.mjs
import { URL, pathToFileURL } from 'url';
import CoffeeScript from 'coffeescript';

const baseURL = pathToFileURL(`${process.cwd()}/`).href;

// CoffeeScript files end in .coffee, .litcoffee or .coffee.md.
const extensionsRegex = /\.coffee$|\.litcoffee$|\.coffee\.md$/;

export function resolve(specifier, context, defaultResolve) {
  const { parentURL = baseURL } = context;

  // Node.js normally errors on unknown file extensions, so return a URL for
  // specifiers ending in the CoffeeScript file extensions.
  if (extensionsRegex.test(specifier)) {
    return {
      url: new URL(specifier, parentURL).href
    };
  }

  // Let Node.js handle all other specifiers.
  return defaultResolve(specifier, context, defaultResolve);
}

export function getFormat(url, context, defaultGetFormat) {
  // Now that we patched resolve to let CoffeeScript URLs through, we need to
  // tell Node.js what format such URLs should be interpreted as. For the
  // purposes of this loader, all CoffeeScript URLs are ES modules.
  if (extensionsRegex.test(url)) {
    return {
      format: 'module'
    };
  }

  // Let Node.js handle all other URLs.
  return defaultGetFormat(url, context, defaultGetFormat);
}

export function transformSource(source, context, defaultTransformSource) {
  const { url, format } = context;

  if (extensionsRegex.test(url)) {
    return {
      source: CoffeeScript.compile(source, { bare: true })
    };
  }

  // Let Node.js handle all other sources.
  return defaultTransformSource(source, context, defaultTransformSource);
}
# main.coffee
import { scream } from './scream.coffee'
console.log scream 'hello, world'

import { version } from 'process'
console.log "Brought to you by Node.js version #{version}"
# scream.coffee
export scream = (str) -> str.toUpperCase()

With the preceding loader, running node --experimental-loader ./coffeescript-loader.mjs main.coffee causes main.coffee to be turned into JavaScript after its source code is loaded from disk but before Node.js executes it; and so on for any .coffee, .litcoffee or .coffee.md files referenced via import statements of any loaded file.

解析算法#

特性#

The resolver has the following properties:

  • FileURL-based resolution as is used by ES modules
  • Support for builtin module loading
  • Relative and absolute URL resolution
  • No default extensions
  • No folder mains
  • Bare specifier package resolution lookup through node_modules

解析器算法#

The algorithm to load an ES module specifier is given through the ESM_RESOLVE method below. It returns the resolved URL for a module specifier relative to a parentURL.

The algorithm to determine the module format of a resolved URL is provided by ESM_FORMAT, which returns the unique module format for any file. The "module" format is returned for an ECMAScript Module, while the "commonjs" format is used to indicate loading through the legacy CommonJS loader. Additional formats such as "addon" can be extended in future updates.

In the following algorithms, all subroutine errors are propagated as errors of these top-level routines unless stated otherwise.

defaultConditions is the conditional environment name array, ["node", "import"].

The resolver can throw the following errors:

  • Invalid Module Specifier: Module specifier is an invalid URL, package name or package subpath specifier.
  • Invalid Package Configuration: package.json configuration is invalid or contains an invalid configuration.
  • Invalid Package Target: Package exports or imports define a target module for the package that is an invalid type or string target.
  • Package Path Not Exported: Package exports do not define or permit a target subpath in the package for the given module.
  • Package Import Not Defined: Package imports do not define the specifier.
  • Module Not Found: The package or module requested does not exist.

解析器算法规范#

ESM_RESOLVE(specifier, parentURL)

  1. Let resolved be undefined.
  2. If specifier is a valid URL, then
    1. Set resolved to the result of parsing and reserializing specifier as a URL.
  3. Otherwise, if specifier starts with "/", "./" or "../", then
    1. Set resolved to the URL resolution of specifier relative to parentURL.
  4. Otherwise, if specifier starts with "#", then
    1. Set resolved to the destructured value of the result of PACKAGE_IMPORTS_RESOLVE(specifier, parentURL, defaultConditions).
  5. Otherwise,
    1. Note: specifier is now a bare specifier.
    2. Set resolved the result of PACKAGE_RESOLVE(specifier, parentURL).
  6. If resolved contains any percent encodings of "/" or "\" ("%2f" and "%5C" respectively), then
    1. Throw an Invalid Module Specifier error.
  7. If the file at resolved is a directory, then
    1. Throw an Unsupported Directory Import error.
  8. If the file at resolved does not exist, then
    1. Throw a Module Not Found error.
  9. Set resolved to the real path of resolved.
  10. Let format be the result of ESM_FORMAT(resolved).
  11. Load resolved as module format, format.
  12. Return resolved.

PACKAGE_RESOLVE(packageSpecifier, parentURL)

  1. Let packageName be undefined.
  2. If packageSpecifier is an empty string, then
    1. Throw an Invalid Module Specifier error.
  3. If packageSpecifier does not start with "@", then
    1. Set packageName to the substring of packageSpecifier until the first "/" separator or the end of the string.
  4. Otherwise,
    1. If packageSpecifier does not contain a "/" separator, then
      1. Throw an Invalid Module Specifier error.
    2. Set packageName to the substring of packageSpecifier until the second "/" separator or the end of the string.
  5. If packageName starts with "." or contains "\" or "%", then
    1. Throw an Invalid Module Specifier error.
  6. Let packageSubpath be "." concatenated with the substring of packageSpecifier from the position at the length of packageName.
  7. Let selfUrl be the result of PACKAGE_SELF_RESOLVE(packageName, packageSubpath, parentURL).
  8. If selfUrl is not undefined, return selfUrl.
  9. If packageSubpath is "." and packageName is a Node.js builtin module, then
    1. Return the string "node:" concatenated with packageSpecifier.
  10. While parentURL is not the file system root,
    1. Let packageURL be the URL resolution of "node_modules/" concatenated with packageSpecifier, relative to parentURL.
    2. Set parentURL to the parent folder URL of parentURL.
    3. If the folder at packageURL does not exist, then
      1. Set parentURL to the parent URL path of parentURL.
      2. Continue the next loop iteration.
    4. Let pjson be the result of READ_PACKAGE_JSON(packageURL).
    5. If pjson is not null and pjson.exports is not null or undefined, then
      1. Let exports be pjson.exports.
      2. Return the resolved destructured value of the result of PACKAGE_EXPORTS_RESOLVE(packageURL, packageSubpath, pjson.exports, defaultConditions).
    6. Otherwise, if packageSubpath is equal to ".", then
      1. Return the result applying the legacy LOAD_AS_DIRECTORY CommonJS resolver to packageURL, throwing a Module Not Found error for no resolution.
    7. Otherwise,
      1. Return the URL resolution of packageSubpath in packageURL.
  11. Throw a Module Not Found error.

PACKAGE_SELF_RESOLVE(packageName, packageSubpath, parentURL)

  1. Let packageURL be the result of READ_PACKAGE_SCOPE(parentURL).
  2. If packageURL is null, then
    1. Return undefined.
  3. Let pjson be the result of READ_PACKAGE_JSON(packageURL).
  4. If pjson is null or if pjson.exports is null or undefined, then
    1. Return undefined.
  5. If pjson.name is equal to packageName, then
    1. Return the resolved destructured value of the result of PACKAGE_EXPORTS_RESOLVE(packageURL, subpath, pjson.exports, defaultConditions).
  6. Otherwise, return undefined.

PACKAGE_EXPORTS_RESOLVE(packageURL, subpath, exports, conditions)

  1. If exports is an Object with both a key starting with "." and a key not starting with ".", throw an Invalid Package Configuration error.
  2. If subpath is equal to ".", then
    1. Let mainExport be undefined.
    2. If exports is a String or Array, or an Object containing no keys starting with ".", then
      1. Set mainExport to exports.
    3. Otherwise if exports is an Object containing a "." property, then
      1. Set mainExport to exports["."].
    4. If mainExport is not undefined, then
      1. Let resolved be the result of PACKAGE_TARGET_RESOLVE( packageURL, mainExport, "", false, false, conditions).
      2. If resolved is not null or undefined, then
        1. Return resolved.
  3. Otherwise, if exports is an Object and all keys of exports start with ".", then
    1. Let matchKey be the string "./" concatenated with subpath.
    2. Let resolvedMatch be result of PACKAGE_IMPORTS_EXPORTS_RESOLVE( matchKey, exports, packageURL, false, conditions).
    3. If resolvedMatch.resolve is not null or undefined, then
      1. Return resolvedMatch.
  4. Throw a Package Path Not Exported error.

PACKAGE_IMPORTS_RESOLVE(specifier, parentURL, conditions)

  1. Assert: specifier begins with "#".
  2. If specifier is exactly equal to "#" or starts with "#/", then
    1. Throw an Invalid Module Specifier error.
  3. Let packageURL be the result of READ_PACKAGE_SCOPE(parentURL).
  4. If packageURL is not null, then
    1. Let pjson be the result of READ_PACKAGE_JSON(packageURL).
    2. If pjson.imports is a non-null Object, then
      1. Let resolvedMatch be the result of PACKAGE_IMPORTS_EXPORTS_RESOLVE(specifier, pjson.imports, packageURL, true, conditions).
      2. If resolvedMatch.resolve is not null or undefined, then
        1. Return resolvedMatch.
  5. Throw a Package Import Not Defined error.

PACKAGE_IMPORTS_EXPORTS_RESOLVE(matchKey, matchObj, packageURL, isImports, conditions)

  1. If matchKey is a key of matchObj, and does not end in "*", then
    1. Let target be the value of matchObj[matchKey].
    2. Let resolved be the result of PACKAGE_TARGET_RESOLVE( packageURL, target, "", false, isImports, conditions).
    3. Return the object { resolved, exact: true }.
  2. Let expansionKeys be the list of keys of matchObj ending in "/" or "*", sorted by length descending.
  3. For each key expansionKey in expansionKeys, do
    1. If expansionKey ends in "*" and matchKey starts with but is not equal to the substring of expansionKey excluding the last "*" character, then
      1. Let target be the value of matchObj[expansionKey].
      2. Let subpath be the substring of matchKey starting at the index of the length of expansionKey minus one.
      3. Let resolved be the result of PACKAGE_TARGET_RESOLVE( packageURL, target, subpath, true, isImports, conditions).
      4. Return the object { resolved, exact: true }.
    2. If matchKey starts with expansionKey, then
      1. Let target be the value of matchObj[expansionKey].
      2. Let subpath be the substring of matchKey starting at the index of the length of expansionKey.
      3. Let resolved be the result of PACKAGE_TARGET_RESOLVE( packageURL, target, subpath, false, isImports, conditions).
      4. Return the object { resolved, exact: false }.
  4. Return the object { resolved: null, exact: true }.

PACKAGE_TARGET_RESOLVE(packageURL, target, subpath, pattern, internal, conditions)

  1. If target is a String, then
    1. If pattern is false, subpath has non-zero length and target does not end with "/", throw an Invalid Module Specifier error.
    2. If target does not start with "./", then
      1. If internal is true and target does not start with "../" or "/" and is not a valid URL, then
        1. If pattern is true, then
          1. Return PACKAGE_RESOLVE(target with every instance of "*" replaced by subpath, packageURL + "/")_.
        2. Return PACKAGE_RESOLVE(target + subpath, packageURL + "/")_.
      2. Otherwise, throw an Invalid Package Target error.
    3. If target split on "/" or "\" contains any ".", ".." or "node_modules" segments after the first segment, throw an Invalid Package Target error.
    4. Let resolvedTarget be the URL resolution of the concatenation of packageURL and target.
    5. Assert: resolvedTarget is contained in packageURL.
    6. If subpath split on "/" or "\" contains any ".", ".." or "node_modules" segments, throw an Invalid Module Specifier error.
    7. If pattern is true, then
      1. Return the URL resolution of resolvedTarget with every instance of "*" replaced with subpath.
    8. Otherwise,
      1. Return the URL resolution of the concatenation of subpath and resolvedTarget.
  2. Otherwise, if target is a non-null Object, then
    1. If exports contains any index property keys, as defined in ECMA-262 6.1.7 Array Index, throw an Invalid Package Configuration error.
    2. For each property p of target, in object insertion order as,
      1. If p equals "default" or conditions contains an entry for p, then
        1. Let targetValue be the value of the p property in target.
        2. Let resolved be the result of PACKAGE_TARGET_RESOLVE( packageURL, targetValue, subpath, pattern, internal, conditions).
        3. If resolved is equal to undefined, continue the loop.
        4. Return resolved.
    3. Return undefined.
  3. Otherwise, if target is an Array, then
    1. If _target.length is zero, return null.
    2. For each item targetValue in target, do
      1. Let resolved be the result of PACKAGE_TARGET_RESOLVE( packageURL, targetValue, subpath, pattern, internal, conditions), continuing the loop on any Invalid Package Target error.
      2. If resolved is undefined, continue the loop.
      3. Return resolved.
    3. Return or throw the last fallback resolution null return or error.
  4. Otherwise, if target is null, return null.
  5. Otherwise throw an Invalid Package Target error.

ESM_FORMAT(url)

  1. Assert: url corresponds to an existing file.
  2. Let pjson be the result of READ_PACKAGE_SCOPE(url).
  3. If url ends in ".mjs", then
    1. Return "module".
  4. If url ends in ".cjs", then
    1. Return "commonjs".
  5. If pjson?.type exists and is "module", then
    1. If url ends in ".js", then
      1. Return "module".
    2. Throw an Unsupported File Extension error.
  6. Otherwise,
    1. Throw an Unsupported File Extension error.

READ_PACKAGE_SCOPE(url)

  1. Let scopeURL be url.
  2. While scopeURL is not the file system root,
    1. Set scopeURL to the parent URL of scopeURL.
    2. If scopeURL ends in a "node_modules" path segment, return null.
    3. Let pjson be the result of READ_PACKAGE_JSON(scopeURL).
    4. If pjson is not null, then
      1. Return pjson.
  3. Return null.

READ_PACKAGE_JSON(packageURL)

  1. Let pjsonURL be the resolution of "package.json" within packageURL.
  2. If the file at pjsonURL does not exist, then
    1. Return null.
  3. If the file at packageURL does not parse as valid JSON, then
    1. Throw an Invalid Package Configuration error.
  4. Return the parsed JSON source of the file at pjsonURL.

自定义的 ESM 说明符解析算法#

The current specifier resolution does not support all default behavior of the CommonJS loader. One of the behavior differences is automatic resolution of file extensions and the ability to import directories that have an index file.

The --experimental-specifier-resolution=[mode] flag can be used to customize the extension resolution algorithm. The default mode is explicit, which requires the full path to a module be provided to the loader. To enable the automatic extension resolution and importing from directories that include an index file use the node mode.

$ node index.mjs
success!
$ node index # Failure!
Error: Cannot find module
$ node --experimental-specifier-resolution=node index
success!