The AutomaticPrefetchPlugin
discovers all modules from the previous compilation upfront while watching for changes, trying to improve the incremental build times. Compared to PrefetchPlugin
which discovers a single module upfront.
webpack.config.js
module.exports = {
// ...
plugins: [new webpack.AutomaticPrefetchPlugin()],
};
Adds a banner to the top of each generated chunk.
const webpack = require('webpack');
new webpack.BannerPlugin(banner);
// or
new webpack.BannerPlugin(options);
{
banner: string | function, // the banner as string or function, it will be wrapped in a comment
raw: boolean, // if true, banner will not be wrapped in a comment
entryOnly: boolean, // if true, the banner will only be added to the entry chunks
test: string | RegExp | [string, RegExp], // Include all modules that pass test assertion.
include: string | RegExp | [string, RegExp], // Include all modules matching any of these conditions.
exclude: string | RegExp | [string, RegExp], // Exclude all modules matching any of these conditions.
}
import webpack from 'webpack';
// string
new webpack.BannerPlugin({
banner: 'hello world',
});
// function
new webpack.BannerPlugin({
banner: (yourVariable) => {
return `yourVariable: ${yourVariable}`;
},
});
Since webpack 2.5.0, placeholders are evaluated in the banner
string:
import webpack from 'webpack';
new webpack.BannerPlugin({
banner:
'fullhash:[fullhash], chunkhash:[chunkhash], name:[name], filebase:[filebase], query:[query], file:[file]',
});
The CommonsChunkPlugin
is an opt-in feature that creates a separate file (known as a chunk), consisting of common modules shared between multiple entry points.
By separating common modules from bundles, the resulting chunked file can be loaded once initially, and stored in cache for later use. This results in page speed optimizations as the browser can quickly serve the shared code from cache, rather than being forced to load a larger bundle whenever a new page is visited.
new webpack.optimize.CommonsChunkPlugin(options);
{
name: string, // or
names: string[],
// The chunk name of the commons chunk. An existing chunk can be selected by passing a name of an existing chunk.
// If an array of strings is passed this is equal to invoking the plugin multiple times for each chunk name.
// If omitted and `options.async` or `options.children` is set all chunks are used, otherwise `options.filename`
// is used as chunk name.
// When using `options.async` to create common chunks from other async chunks you must specify an entry-point
// chunk name here instead of omitting the `option.name`.
filename: string,
// The filename template for the commons chunk. Can contain the same placeholders as `output.filename`.
// If omitted the original filename is not modified (usually `output.filename` or `output.chunkFilename`).
// This option is not permitted if you're using `options.async` as well, see below for more details.
minChunks: number|Infinity|function(module, count) => boolean,
// The minimum number of chunks which need to contain a module before it's moved into the commons chunk.
// The number must be greater than or equal 2 and lower than or equal to the number of chunks.
// Passing `Infinity` creates the commons chunk, but moves no modules into it.
// By providing a `function` you can add custom logic. (Defaults to the number of chunks)
chunks: string[],
// Select the source chunks by chunk names. The chunk must be a child of the commons chunk.
// If omitted all entry chunks are selected.
children: boolean,
// If `true` all children of the commons chunk are selected
deepChildren: boolean,
// If `true` all descendants of the commons chunk are selected
async: boolean|string,
// If `true` a new async commons chunk is created as child of `options.name` and sibling of `options.chunks`.
// It is loaded in parallel with `options.chunks`.
// Instead of using `option.filename`, it is possible to change the name of the output file by providing
// the desired string here instead of `true`.
minSize: number,
// Minimum size of all common module before a commons chunk is created.
}
Generate an extra chunk, which contains common modules shared between entry points.
new webpack.optimize.CommonsChunkPlugin({
name: 'commons',
// (the commons chunk name)
filename: 'commons.js',
// (the filename of the commons chunk)
// minChunks: 3,
// (Modules must be shared between 3 entries)
// chunks: ["pageA", "pageB"],
// (Only use these entries)
});
You must load the generated chunk before the entry point:
<script src="commons.js" charset="utf-8"></script>
<script src="entry.bundle.js" charset="utf-8"></script>
Split your code into vendor and application.
module.exports = {
//...
entry: {
vendor: ['jquery', 'other-lib'],
app: './entry',
},
plugins: [
new webpack.optimize.CommonsChunkPlugin({
name: 'vendor',
// filename: "vendor.js"
// (Give the chunk a different name)
minChunks: Infinity,
// (with more entries, this ensures that no other module
// goes into the vendor chunk)
}),
],
};
<script src="vendor.js" charset="utf-8"></script>
<script src="app.js" charset="utf-8"></script>
With Code Splitting, multiple child chunks of an entry chunk can have common dependencies. To prevent duplication these can be moved into the parent. This reduces overall size, but does have a negative effect on the initial load time. If it is expected that users will need to download many sibling chunks, i.e. children of the entry chunk, then this should improve load time overall.
new webpack.optimize.CommonsChunkPlugin({
// names: ["app", "subPageA"]
// (choose the chunks, or omit for all chunks)
children: true,
// (select all children of chosen chunks)
// minChunks: 3,
// (3 children must share the module before it's moved)
});
Similar to the above one, but instead of moving common modules into the parent (which increases initial load time) a new async-loaded additional commons chunk is used. This is automatically downloaded in parallel when the additional chunk is downloaded.
new webpack.optimize.CommonsChunkPlugin({
name: 'app',
// or
names: ['app', 'subPageA'],
// the name or list of names must match the name or names
// of the entry points that create the async chunks
children: true,
// (use all children of the chunk)
async: true,
// (create an async commons chunk)
minChunks: 3,
// (3 children must share the module before it's separated)
});
minChunks
property a functionYou also have the ability to pass the minChunks
property a function. This function is called by the CommonsChunkPlugin
and calls the function with module
and count
arguments.
The module
argument represents each module in the chunks you have provided via the name
/names
property.
module
has the shape of a NormalModule, which has two particularly useful properties for this use case:
module.context
: The directory that stores the file. For example: '/my_project/node_modules/example-dependency'
module.resource
: The name of the file being processed. For example: '/my_project/node_modules/example-dependency/index.js'
The count
argument represents how many chunks the module
is used in.
This option is useful when you want to have fine-grained control over how the CommonsChunk algorithm determines where modules should be moved to.
new webpack.optimize.CommonsChunkPlugin({
name: 'my-single-lib-chunk',
filename: 'my-single-lib-chunk.js',
minChunks: function (module, count) {
// If module has a path, and inside of the path exists the name "somelib",
// and it is used in 3 separate chunks/entries, then break it out into
// a separate chunk with chunk keyname "my-single-lib-chunk", and filename "my-single-lib-chunk.js"
return module.resource && /somelib/.test(module.resource) && count === 3;
},
});
As seen above, this example allows you to move only one lib to a separate file if and only if all conditions are met inside the function.
This concept may be used to obtain implicit common vendor chunks:
new webpack.optimize.CommonsChunkPlugin({
name: 'vendor',
minChunks: function (module) {
// this assumes your vendor imports exist in the node_modules directory
return module.context && module.context.includes('node_modules');
},
});
To extract the webpack bootstrap logic into a separate file, use the CommonsChunkPlugin
on a name
which is not defined as entry
. Commonly the name manifest
is used. See the caching guide for details.
new webpack.optimize.CommonsChunkPlugin({
name: 'manifest',
minChunks: Infinity,
});
Since the vendor
and manifest
chunk use a different definition for minChunks
, you need to invoke the plugin twice:
[
new webpack.optimize.CommonsChunkPlugin({
name: 'vendor',
minChunks: function (module) {
return module.context && module.context.includes('node_modules');
},
}),
new webpack.optimize.CommonsChunkPlugin({
name: 'manifest',
minChunks: Infinity,
}),
];
Context refers to a require with an expression such as require('./locale/' + name + '.json')
.
The ContextExclusionPlugin
allows you to exclude context. Provide RegExp as an argument when initializing the Plugin to exclude all context that matches it.
webpack.config.js
module.exports = {
plugins: [new webpack.ContextExclusionPlugin(/dont/)],
};
Context refers to a require with an expression such as require('./locale/' + name + '.json')
. When encountering such an expression, webpack infers the directory ('./locale/'
) and a regular expression (/^.*\.json$/
). Since the name
is not known at compile time, webpack includes every file as module in the bundle.
The ContextReplacementPlugin
allows you to override the inferred information. There are various ways to configure the plugin:
new webpack.ContextReplacementPlugin(
resourceRegExp: RegExp,
newContentResource?: string,
newContentRecursive?: boolean,
newContentRegExp?: RegExp
)
If the resource (directory) matches resourceRegExp
, the plugin replaces the default resource, recursive flag or generated regular expression with newContentResource
, newContentRecursive
or newContextRegExp
respectively. If newContentResource
is relative, it is resolved relative to the previous resource.
Here's a small example to restrict module usage:
new webpack.ContextReplacementPlugin(/moment[/\\]locale$/, /de|fr|hu/);
The moment/locale
context is restricted to files matching /de|fr|hu/
. Thus only those locales are included (see this issue for more information).
new webpack.ContextReplacementPlugin(
resourceRegExp: RegExp,
newContentCallback: (data) => void
);
The newContentCallback
function is given a data
object of the ContextModuleFactory
and is expected to overwrite the request
attribute of the supplied object.
Using this callback we can dynamically redirect requests to a new location:
new webpack.ContextReplacementPlugin(/^\.\/locale$/, (context) => {
if (!/\/moment\//.test(context.context)) return;
Object.assign(context, {
regExp: /^\.\/\w+/,
request: '../../locale', // resolved relatively
});
});
The newContentResource
and newContentCreateContextMap
parameters are also available:
new webpack.ContextReplacementPlugin(
resourceRegExp: RegExp,
newContentResource: string,
newContentCreateContextMap: object // mapping runtime-request (userRequest) to compile-time-request (request)
);
These two parameters can be used together to redirect requests in a more targeted way. The newContentCreateContextMap
allows you to map runtime requests to compile requests in the form of an object:
new ContextReplacementPlugin(/selector/, './folder', {
'./request': './request',
'./other-request': './new-request',
});
The DefinePlugin
replaces variables in your code with other values or expressions at compile time. This can be useful for allowing different behavior between development builds and production builds. If you perform logging in your development build but not in the production build you might use a global constant to determine whether logging takes place. That's where DefinePlugin
shines, set it and forget it rules for development and production builds.
new webpack.DefinePlugin({
// Definitions...
});
Each key passed into DefinePlugin
is an identifier or multiple identifiers joined with .
.
typeof
to the key, it's only defined for typeof calls.The values will be inlined into the code allowing a minification pass to remove the redundant conditional.
new webpack.DefinePlugin({
PRODUCTION: JSON.stringify(true),
VERSION: JSON.stringify('5fa3b9'),
BROWSER_SUPPORTS_HTML5: true,
TWO: '1+1',
'typeof window': JSON.stringify('object'),
'process.env.NODE_ENV': JSON.stringify(process.env.NODE_ENV),
});
console.log('Running App version ' + VERSION);
if (!BROWSER_SUPPORTS_HTML5) require('html5shiv');
if (!PRODUCTION) {
console.log('Debug info');
}
if (PRODUCTION) {
console.log('Production log');
}
After passing through webpack with no minification results in:
if (!true) {
console.log('Debug info');
}
if (true) {
console.log('Production log');
}
and then after a minification pass results in:
console.log('Production log');
Enable/disable features in production/development build using feature flags.
new webpack.DefinePlugin({
NICE_FEATURE: JSON.stringify(true),
EXPERIMENTAL_FEATURE: JSON.stringify(false),
});
Use a different service URL in production/development builds:
new webpack.DefinePlugin({
SERVICE_URL: JSON.stringify('https://dev.example.com'),
});
runtimeValue
function (getterFunction, [string] | true | object) => getterFunction()
It is possible to define variables with values that rely on files and will be re-evaluated when such files change in the file system. This means webpack will rebuild when such watched files change.
There're two arguments for webpack.DefinePlugin.runtimeValue
function:
The first argument is a function(module, key, version)
that should return the value to be assigned to the definition.
The second argument could either be an array of file paths to watch for or a true
to flag the module as uncacheable. Since 5.26.0, it can also take an object argument with the following properties:
fileDependencies?: string[]
A list of files the function depends on.contextDependencies?: string[]
A list of directories the function depends on.missingDependencies?: string[]
A list of not existing files the function depends on.buildDependencies?: string[]
A list of build dependencies the function depends on.version?: string | () => string
A version of the function.const fileDep = path.resolve(__dirname, 'sample.txt');
new webpack.DefinePlugin({
BUILT_AT: webpack.DefinePlugin.runtimeValue(Date.now, {
fileDependencies: [fileDep],
}),
});
The value of BUILT_AT
would be the time at which the 'sample.txt'
was last updated in the file system, e.g. 1597953013291
.
The DllPlugin
and DllReferencePlugin
provide means to split bundles in a way that can drastically improve build time performance. The term "DLL" stands for Dynamic-link library which was originally introduced by Microsoft.
DllPlugin
This plugin is used in a separate webpack configuration exclusively to create a dll-only-bundle. It creates a manifest.json
file, which is used by the DllReferencePlugin
to map dependencies.
context
(optional): context of requests in the manifest file (defaults to the webpack context.)format
(boolean = false): If true
, manifest json file (output) will be formatted.name
: name of the exposed dll function (TemplatePaths: [fullhash]
& [name]
)path
: absolute path to the manifest json file (output)entryOnly
(boolean = true): if true
, only entry points will be exposedtype
: type of the dll bundlenew webpack.DllPlugin(options);
Creates a manifest.json
which is written to the given path
. It contains mappings from require and import requests to module ids. It is used by the DllReferencePlugin
.
Combine this plugin with output.library
option to expose (aka, put into the global scope) the dll function.
DllReferencePlugin
This plugin is used in the primary webpack config, it references the dll-only-bundle(s) to require pre-built dependencies.
context
: (absolute path) context of requests in the manifest (or content property)extensions
: Extensions used to resolve modules in the dll bundle (only used when using 'scope').manifest
: an object containing content
and name
or a string to the absolute path of the JSON manifest to be loaded upon compilationcontent
(optional): the mappings from request to module id (defaults to manifest.content
)name
(optional): an identifier where the dll is exposed (defaults to manifest.name
) (see also externals
)scope
(optional): prefix which is used for accessing the content of the dllsourceType
(optional): how the dll is exposed (libraryTarget)new webpack.DllReferencePlugin(options);
References a dll manifest file to map dependency names to module ids, then requires them as needed using the internal __webpack_require__
function.
This plugin can be used in two different modes, scoped and mapped.
The content of the dll is accessible under a module prefix. i.e. with scope = 'xyz'
a file abc
in the dll can be access via require('xyz/abc')
.
The content of the dll is mapped to the current directory. If a required file matches a file in the dll (after resolving), then the file from the dll is used instead.
Because this happens after resolving every file in the dll bundle, the same paths must be available for the consumer of the dll bundle. i.e. if the dll contains lodash
and the file abc
, require('lodash')
and require('./abc')
will be used from the dll, rather than building them into the main bundle.
webpack.vendor.config.js
const path = require('path');
new webpack.DllPlugin({
context: __dirname,
name: '[name]_[fullhash]',
path: path.join(__dirname, 'manifest.json'),
});
webpack.app.config.js
new webpack.DllReferencePlugin({
context: __dirname,
manifest: require('./manifest.json'),
scope: 'xyz',
sourceType: 'commonjs2',
});
Two separate example folders. Demonstrates scope and context.
The EnvironmentPlugin
is shorthand for using the DefinePlugin
on process.env
keys.
The EnvironmentPlugin
accepts either an array of keys or an object mapping its keys to their default values.
new webpack.EnvironmentPlugin(['NODE_ENV', 'DEBUG']);
This is equivalent to the following DefinePlugin
application:
new webpack.DefinePlugin({
'process.env.NODE_ENV': JSON.stringify(process.env.NODE_ENV),
'process.env.DEBUG': JSON.stringify(process.env.DEBUG),
});
Alternatively, the EnvironmentPlugin
supports an object, which maps keys to their default values. The default value for a key is taken if the key is undefined in process.env
.
new webpack.EnvironmentPlugin({
NODE_ENV: 'development', // use 'development' unless process.env.NODE_ENV is defined
DEBUG: false,
});
Example:
Let's investigate the result when running the previous EnvironmentPlugin
configuration on a test file entry.js
:
if (process.env.NODE_ENV === 'production') {
console.log('Welcome to production');
}
if (process.env.DEBUG) {
console.log('Debugging output');
}
When executing NODE_ENV=production webpack
in the terminal to build, entry.js
becomes this:
if ('production' === 'production') {
// <-- 'production' from NODE_ENV is taken
console.log('Welcome to production');
}
if (false) {
// <-- default value is taken
console.log('Debugging output');
}
Running DEBUG=false webpack
yields:
if ('development' === 'production') {
// <-- default value is taken
console.log('Welcome to production');
}
if ('false') {
// <-- 'false' from DEBUG is taken
console.log('Debugging output');
}
The following EnvironmentPlugin
configuration provides process.env.GIT_VERSION
(such as "v5.4.0-2-g25139f57f") and process.env.GIT_AUTHOR_DATE
(such as "2020-11-04T12:25:16+01:00") corresponding to the last Git commit of the repository:
const child_process = require('child_process');
function git(command) {
return child_process.execSync(`git ${command}`, { encoding: 'utf8' }).trim();
}
new webpack.EnvironmentPlugin({
GIT_VERSION: git('describe --always'),
GIT_AUTHOR_DATE: git('log -1 --format=%aI'),
});
DotenvPlugin
The third-party DotenvPlugin
(dotenv-webpack
) allows you to expose (a subset of) dotenv variables:
// .env
DB_HOST=127.0.0.1
DB_PASS=foobar
S3_API=mysecretkey
new Dotenv({
path: './.env', // Path to .env file (this is the default)
safe: true, // load .env.example (defaults to "false" which does not use dotenv-safe)
});
This plugin enables more fine grained control of source map generation. It is also enabled automatically by certain settings of the devtool
configuration option.
new webpack.EvalSourceMapDevToolPlugin(options);
The following options are supported:
test
(string|RegExp|array
): Include source maps for modules based on their extension (defaults to .js
and .css
).include
(string|RegExp|array
): Include source maps for module paths that match the given value.exclude
(string|RegExp|array
): Exclude modules that match the given value from source map generation.append
(string
): Appends the given value to the original asset. Usually the #sourceMappingURL
comment. [url]
is replaced with a URL to the source map file. false
disables the appending.moduleFilenameTemplate
(string
): See output.devtoolModuleFilenameTemplate
.module
(boolean
): Indicates whether loaders should generate source maps (defaults to true
).columns
(boolean
): Indicates whether column mappings should be used (defaults to true
).protocol
(string
): Allows user to override default protocol (webpack-internal://
)The following examples demonstrate some common use cases for this plugin.
You can use the following code to replace the configuration option devtool: eval-source-map
with an equivalent custom plugin configuration:
module.exports = {
// ...
devtool: false,
plugins: [new webpack.EvalSourceMapDevToolPlugin({})],
};
The following code would exclude source maps for any modules in the vendor.js
bundle:
new webpack.EvalSourceMapDevToolPlugin({
exclude: ['vendor.js'],
});
This plugin will cause hashes to be based on the relative path of the module, generating a four character string as the module id. Suggested for use in production.
new webpack.ids.HashedModuleIdsPlugin({
// Options...
});
This plugin supports the following options:
context
: The context directory (absolute path) for creating names.hashFunction
: The hashing algorithm to use, defaults to 'md4'
. All functions from Node.JS' crypto.createHash
are supported.hashDigest
: The encoding to use when generating the hash, defaults to 'base64'
. All encodings from Node.JS' hash.digest
are supported.hashDigestLength
: The prefix length of the hash digest to use, defaults to 4
. Note that some generated ids might be longer than specified here, to avoid module id collisions.Here's an example of how this plugin might be used:
new webpack.ids.HashedModuleIdsPlugin({
context: __dirname,
hashFunction: 'sha256',
hashDigest: 'hex',
hashDigestLength: 20,
});
Enables Hot Module Replacement, otherwise known as HMR.
Enabling HMR is straightforward and in most cases no options are necessary.
new webpack.HotModuleReplacementPlugin({
// Options...
});
The HtmlWebpackPlugin
simplifies creation of HTML files to serve your webpack bundles. This is especially useful for webpack bundles that include a hash in the filename which changes every compilation. You can either let the plugin generate an HTML file for you, supply your own template using lodash templates, or use your own loader.
npm install --save-dev html-webpack-plugin
The plugin will generate an HTML5 file for you that includes all your webpack
bundles in the body using script
tags. Add the plugin to your webpack
configuration as follows:
const HtmlWebpackPlugin = require('html-webpack-plugin');
const path = require('path');
module.exports = {
entry: 'index.js',
output: {
path: path.resolve(__dirname, './dist'),
filename: 'index_bundle.js',
},
plugins: [new HtmlWebpackPlugin()],
};
This will generate a file dist/index.html
containing the following:
<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8" />
<title>webpack App</title>
</head>
<body>
<script src="index_bundle.js"></script>
</body>
</html>
If you have multiple webpack entry points, they will all be included with <script>
tags in the generated HTML.
If you have any CSS assets in webpack's output (for example, CSS extracted with the MiniCssExtractPlugin) then these will be included with <link>
tags in the <head>
element of generated HTML.
For all configuration options, please see the plugin documentation.
The plugin supports addons. For a list see the documentation.
IgnorePlugin prevents the generation of modules for import
or require
calls matching the regular expressions or filter functions:
resourceRegExp
: A RegExp to test the resource against.contextRegExp
: (optional) A RegExp to test the context (directory) against.new webpack.IgnorePlugin({ resourceRegExp, contextRegExp });
// Supported in webpack 4 and earlier, unsupported in webpack 5:
new webpack.IgnorePlugin(resourceRegExp, [contextRegExp]);
checkResource (resource, context)
A Filter function that receives resource
and context
as arguments, must return boolean.checkContext (context)
was removed in webpack 5 as checkResource
already gets context.new webpack.IgnorePlugin({
checkResource(resource) {
// do something with resource
return true | false;
},
});
As of moment 2.18, all locales are bundled together with the core library (see this GitHub issue).
The resourceRegExp
parameter passed to IgnorePlugin
is not tested against the resolved file names or absolute module names being imported or required, but rather against the string passed to require
or import
within the source code where the import is taking place. For example, if you're trying to exclude node_modules/moment/locale/*.js
, this won't work:
-new webpack.IgnorePlugin({ resourceRegExp: /moment\/locale\// });
Rather, because moment
imports with this code:
require('./locale/' + name);
...your first regexp must match that './locale/'
string. The second contextRegExp
parameter is then used to select specific directories from where the import took place. The following will cause those locale files to be ignored:
new webpack.IgnorePlugin({
resourceRegExp: /^\.\/locale$/,
contextRegExp: /moment$/,
});
...which means "any require statement matching './locale'
from any directories ending with 'moment'
will be ignored.
This is a list of plugins which are used by webpack internally.
Categories of internal plugins:
Plugins affecting the environment of the compiler.
webpack.node.NodeEnvironmentPlugin()
Applies Node.js style filesystem to the compiler.
Plugins affecting the compiler
MemoryCachePlugin()
Adds a cache to the compiler, where modules are cached in memory.
ProgressPlugin(handler)
Hook into the compiler to extract progress information. The handler
must have the signature function(percentage, message)
. Percentage is called with a value between 0 and 1, where 0 indicates the start and 1 the end.
RecordIdsPlugin()
Saves and restores module and chunk ids from records.
Plugins, which add entry chunks to the compilation.
EntryPlugin(context, request, chunkName)
Adds an entry chunk on compilation. The chunk is named chunkName
and contains only one module (plus dependencies). The module is resolved from request
in context
(absolute path).
PrefetchPlugin(context, request)
Prefetches request
and dependencies to enable a more parallel compilation. It doesn't create any chunk. The module is resolved from request
in context
(absolute path).
JsonpTemplatePlugin(options)
Chunks are wrapped into JSONP-calls. A loading algorithm is included in entry chunks. It loads chunks by adding a <script>
tag.
options
are the output options.
options.jsonpFunction
is the JSONP function.
options.publicPath
is used as path for loading the chunks.
options.chunkFilename
is the filename under that chunks are expected.
node/NodeTemplatePlugin(options)
Chunks are wrapped into Node.js modules exporting the bundled modules. The entry chunks loads chunks by requiring them.
options
are the output options.
options.chunkFilename
is the filename under that chunks are expected.
LibraryTemplatePlugin(name, target)
The entries chunks are decorated to form a library name
of type type
.
webworker/WebWorkerTemplatePlugin(options)
Chunks are loaded by importScripts
. Else it's similar to JsonpTemplatePlugin
.
options
are the output options.
Decorates the module template by wrapping each module in a eval
annotated with // @sourceURL
.
SourceMapDevToolPlugin(sourceMapFilename, sourceMappingURLComment, moduleFilenameTemplate, fallbackModuleFilenameTemplate)
Decorates the templates by generating a SourceMap for each chunk.
sourceMapFilename
the filename template of the SourceMap. [hash]
, [name]
, [id]
, [file]
and [filebase]
are replaced. If this argument is missing, the SourceMap will be inlined as DataUrl.
HotModuleReplacementPlugin(options)
Add support for hot module replacement. Decorates the templates to add runtime code. Adds module.hot
API.
options.hotUpdateChunkFilename
the filename for hot update chunks.
options.hotUpdateMainFilename
the filename for the hot update manifest.
options.hotUpdateFunction
JSON function name for the hot update.
Plugins affecting the source code of modules.
Make webpack_public_path, webpack_require, webpack_modules and webpack_chunk_load accessible. Ensures that require.valueOf
and require.onError
are not processed by other plugins.
Currently useless. Ensures compatibility with other module loaders.
Tries to evaluate expressions in if (...)
statements and ternaries to replace them with true
/false
for further possible dead branch elimination using hooks fired by the parser.
There are multiple optimizations in production mode regarding dead branches:
webpack will try to evaluate conditional statements. If it succeeds then the dead branch is removed. webpack can't do constant folding unless the compiler knows it. For example:
import { calculateTax } from './tax';
const FOO = 1;
if (FOO === 0) {
// dead branch
calculateTax();
}
In the above example, webpack is unable to prune the branch, but Terser does. However, if FOO
is defined using DefinePlugin, webpack will succeed.
It is important to mention that import { calculateTax } from './tax';
will also get pruned because calculateTax()
call was in the dead branch and got eliminated.
ProvidePlugin(name, request)
If name
is used in a module it is filled by a module loaded by require(<request>)
.
NodeStuffPlugin(options, context)
Provide stuff that is normally available in Node.js modules.
It also ensures that module
is filled with some Node.js stuff if you use it.
Provide stuff that is normally available in require.js.
require[js].config
is removed. require.version
is 0.0.0
. requirejs.onError
is mapped to require.onError
.
node/NodeSourcePlugin(options)
This module adds stuff from Node.js that is not available in non Node.js environments.
It adds polyfills for process
, console
, Buffer
and global
if used. It also binds the built in Node.js replacement modules.
node/NodeTargetPlugin()
The plugins should be used if you run the bundle in a Node.js environment.
If ensures that native modules are loaded correctly even if bundled.
dependencies/AMDPlugin(options)
Provides AMD-style define
and require
to modules. Also bind require.amd
, define.amd
and webpack_amd_options## to the options
passed as parameter.
dependencies/CommonJsPlugin
Provides CommonJs-style require
to modules.
dependencies/RequireContextPlugin(modulesDirectories, extensions)
Provides require.context
. The parameter modulesDirectories
and extensions
are used to find alternative requests for files. It's useful to provide the same arrays as you provide to the resolver.
dependencies/RequireEnsurePlugin()
Provides require.ensure
.
dependencies/RequireIncludePlugin()
Provides require.include
.
DefinePlugin(definitions)
Define constants for identifier.
definitions
is an object.
Note that all plugins under webpack.optimize
namespace should only be used when mode
set to 'none'
. Otherwise you might get into trouble where plugins are applied twice.
optimize/LimitChunkCountPlugin(options)
Merge chunks limit chunk count is lower than options.maxChunks
.
The overhead for each chunks is provided by options.chunkOverhead
or defaults to 10000. Entry chunks sizes are multiplied by options.entryChunkMultiplicator
(or 10).
Chunks that reduce the total size the most are merged first. If multiple combinations are equal the minimal merged size wins.
optimize/MergeDuplicateChunksPlugin()
Chunks with the same modules are merged.
optimize/RemoveEmptyChunksPlugin()
Modules that are included in every parent chunk are removed from the chunk.
optimize/MinChunkSizePlugin(minChunkSize)
Merges chunks until each chunk has the minimum size of minChunkSize
.
See the ModuleConcatenationPlugin page for details.
optimize/FlagIncludedChunksPlugin()
Adds chunk ids of chunks which are included in the chunk. This eliminates unnecessary chunk loads.
optimize/RealContentHashPlugin()
When optimization.realContentHash
option is enabled, webpack will apply RealContentHashPlugin
to compiler internally.
RealContentHashPlugin
provides a updateHash
5.8.0+ hook for customizing hash updating:
const webpack = require('webpack');
const RealContentHashPlugin = webpack.optimize.RealContentHashPlugin;
// ...
compiler.hooks.compilation.tap('MyPlugin', (compilation) => {
const hooks = RealContentHashPlugin.getCompilationHooks(compilation);
hooks.updateHash.tap('MyPlugin', (content, oldHash) => {
// you can calculate the hash here as you wish
});
});
webpack has a rich plugin interface. Most of the features within webpack itself use this plugin interface. This makes webpack flexible.
Name | Description |
---|---|
BannerPlugin | Add a banner to the top of each generated chunk |
CommonsChunkPlugin | Extract common modules shared between chunks |
CompressionWebpackPlugin | Prepare compressed versions of assets to serve them with Content-Encoding |
ContextReplacementPlugin | Override the inferred context of a require expression |
CopyWebpackPlugin | Copies individual files or entire directories to the build directory |
DefinePlugin | Allow global constants configured at compile time |
DllPlugin | Split bundles in order to drastically improve build time |
EnvironmentPlugin | Shorthand for using the DefinePlugin on process.env keys |
EslintWebpackPlugin | A ESLint plugin for webpack |
HotModuleReplacementPlugin | Enable Hot Module Replacement (HMR) |
HtmlWebpackPlugin | Easily create HTML files to serve your bundles |
IgnorePlugin | Exclude certain modules from bundles |
LimitChunkCountPlugin | Set min/max limits for chunking to better control chunking |
MinChunkSizePlugin | Keep chunk size above the specified limit |
MiniCssExtractPlugin | creates a CSS file per JS file which requires CSS |
NoEmitOnErrorsPlugin | Skip the emitting phase when there are compilation errors |
NormalModuleReplacementPlugin | Replace resource(s) that matches a regexp |
NpmInstallWebpackPlugin | Auto-install missing dependencies during development |
ProgressPlugin | Report compilation progress |
ProvidePlugin | Use modules without having to use import/require |
SourceMapDevToolPlugin | Enables a more fine grained control of source maps |
EvalSourceMapDevToolPlugin | Enables a more fine grained control of eval source maps |
TerserPlugin | Uses Terser to minify the JS in your project |
For more third-party plugins, see the list from awesome-webpack.
While writing your code, you may have already added many code split points to load stuff on demand. After compiling you might notice that some chunks are too small - creating larger HTTP overhead. LimitChunkCountPlugin
can post-process your chunks by merging them.
new webpack.optimize.LimitChunkCountPlugin({
// Options...
});
The following options are supported:
maxChunks
number
Limit the maximum number of chunks using a value greater than or equal to 1
. Using 1
will prevent any additional chunks from being added as the entry/main chunk is also included in the count.
webpack.config.js
const webpack = require('webpack');
module.exports = {
// ...
plugins: [
new webpack.optimize.LimitChunkCountPlugin({
maxChunks: 5,
}),
],
};
minChunkSize
Keeping chunk size above the specified limit is no longer a feature of this plugin. Use MinChunkSizePlugin instead.
This plugin and it's options can also be invoked via the CLI:
webpack --optimize-max-chunks 15
Keep chunk size above the specified limit by merging chunks that are smaller than the minChunkSize
.
new webpack.optimize.MinChunkSizePlugin({
minChunkSize: 10000, // Minimum number of characters
});
This plugin and it's options can also be invoked via the CLI:
webpack --optimize-min-chunk-size 10000
In the past, one of webpack’s trade-offs when bundling was that each module in your bundle would be wrapped in individual function closures. These wrapper functions made it slower for your JavaScript to execute in the browser. In comparison, tools like Closure Compiler and RollupJS ‘hoist’ or concatenate the scope of all your modules into one closure and allow for your code to have a faster execution time in the browser.
This plugin will enable the same concatenation behavior in webpack. By default this plugin is already enabled in production mode
and disabled otherwise. If you need to override the production mode
optimization, set the optimization.concatenateModules
option to false
. To enable concatenation behavior in other modes, you can add ModuleConcatenationPlugin
manually or use the optimization.concatenateModules
option:
new webpack.optimize.ModuleConcatenationPlugin();
This concatenation behavior is called “scope hoisting.”
Scope hoisting is specifically a feature made possible by ECMAScript Module syntax. Because of this webpack may fallback to normal bundling based on what kind of modules you are using, and other conditions.
As the article explains, webpack attempts to achieve partial scope hoisting. It will merge modules into a single scope but cannot do so in every case. If webpack cannot merge a module, the two alternatives are Prevent and Root. Prevent means the module must be in its own scope. Root means a new module group will be created. The following conditions determine the outcome:
Condition | Outcome |
---|---|
Non ES6 Module | Prevent |
Imported By Non Import | Root |
Imported From Other Chunk | Root |
Imported By Multiple Other Module Groups | Root |
Imported With import() | Root |
Affected By ProvidePlugin Or Using module | Prevent |
HMR Accepted | Root |
Using eval() | Prevent |
In Multiple Chunks | Prevent |
export * from "cjs-module" | Prevent |
The following pseudo JavaScript explains the algorithm:
modules.forEach((module) => {
const group = new ModuleGroup({
root: module,
});
module.dependencies.forEach((dependency) => {
tryToAdd(group, dependency);
});
if (group.modules.length > 1) {
orderedModules = topologicalSort(group.modules);
concatenatedModule = new ConcatenatedModule(orderedModules);
chunk.add(concatenatedModule);
orderedModules.forEach((groupModule) => {
chunk.remove(groupModule);
});
}
});
function tryToAdd(group, module) {
if (group.has(module)) {
return true;
}
if (!hasPreconditions(module)) {
return false;
}
const nextGroup = group;
const result = module.dependents.reduce((check, dependent) => {
return check && tryToAdd(nextGroup, dependent);
}, true);
if (!result) {
return false;
}
module.dependencies.forEach((dependency) => {
tryToAdd(group, dependency);
});
group.merge(nextGroup);
return true;
}
When using the webpack CLI, the --display-optimization-bailout
flag will display bailout reasons. When using the webpack config, add the following to the stats
object:
module.exports = {
//...
stats: {
// Examine all modules
maxModules: Infinity,
// Display bailout reasons
optimizationBailout: true,
},
};
The ModuleFederationPlugin
allows a build to provide or consume modules with other independent builds at runtime.
const { ModuleFederationPlugin } = require('webpack').container;
module.exports = {
plugins: [
new ModuleFederationPlugin({
// options' typings in typescript
runtime: string | false,
}),
],
};
Create a new runtime chunk with the specified name.
webpack.js.org
const { ModuleFederationPlugin } = require('webpack').container;
module.exports = {
plugins: [
new ModuleFederationPlugin({
runtime: 'my-runtime-name',
}),
],
};
There are three ways to specify the versions of shared libraries.
This syntax allows you to share libraries with package name only. This approach is good for prototyping, but it will not allow you to scale to large production environment given that libraries like react
and react-dom
will require additional requirements.
const { ModuleFederationPlugin } = require('webpack').container;
module.exports = {
plugins: [
new ModuleFederationPlugin({
// adds lodash as shared module
// version is inferred from package.json
// there is no version check for the required version
// so it will always use the higher version found
shared: ['lodash'],
}),
],
};
This syntax provides you more control over each shared library in which you can define package name as the key and version (semver) as the value.
module.exports = {
plugins: [
new ModuleFederationPlugin({
shared: {
// adds lodash as shared module
// version is inferred from package.json
// it will use the highest lodash version that is >= 4.17 and < 5
lodash: '^4.17.0',
},
}),
],
};
This syntax allows you to provide additional hints to each shared package where you define the package name as the key, and the value as an object containing hints to modify sharing behavior.
const deps = require('./package.json').dependencies;
module.exports = {
plugins: [
new ModuleFederationPlugin({
shared: {
// adds react as shared module
react: {
requiredVersion: deps.react,
singleton: true,
},
},
}),
],
};
eager
boolean
This hint will allow webpack to include the provided and fallback module directly instead of fetching the library via an asynchronous request. In other words, this allows to use this shared module in the initial chunk. Also, be careful that all provided and fallback modules will always be downloaded when this hint is enabled.
import
false | string
The provided module that should be placed in the shared scope. This provided module also acts as fallback module if no shared module is found in the shared scope or version isn't valid. (The value for this hint defaults to the property name.)
packageName
string
The package name that is used to determine required version from description file. This is only needed when the package name can't be automatically determined from request.
requiredVersion
false | string
The required version of the package.
shareKey
string
The requested shared module is looked up under this key from the shared scope.
shareScope
string
The name of the shared scope.
singleton
boolean
This hint only allows a single version of the shared module in the shared scope (disabled by default). Some libraries use a global internal state (e.g. react, react-dom). Thus, it is critical to have only one instance of the library running at a time.
strictVersion
boolean
This hint allows webpack to reject the shared module if version is not valid (defaults to true
when local fallback module is available and shared module is not a singleton, otherwise false
, it has no effect if there is no required version specified).
version
false | string
The version of the provided module. It allows webpack to replace lower matching versions, but not higher.
module.exports = {
plugins: [
new ModuleFederationPlugin({
// adds vue as shared module
// version is inferred from package.json
// it will always use the shared version, but print a warning when the shared vue is < 2.6.5 or >= 3
shared: {
vue: {
requiredVersion: '^2.6.5',
singleton: true,
},
},
}),
],
};
module.exports = {
plugins: [
new ModuleFederationPlugin({
// adds vue as shared module
// there is no local version provided
// it will emit a warning if the shared vue is < 2.6.5 or >= 3
shared: {
vue: {
import: false,
requiredVersion: '^2.6.5',
},
},
}),
],
};
module.exports = {
plugins: [
new ModuleFederationPlugin({
// adds vue as shared module
// there is no local version provided
// it will throw an error when the shared vue is < 2.6.5 or >= 3
shared: {
vue: {
import: false,
requiredVersion: '^2.6.5',
strictVersion: true,
},
},
}),
],
};
module.exports = {
plugins: [
new ModuleFederationPlugin({
shared: {
'my-vue': {
// can be referenced by import "my-vue"
import: 'vue', // the "vue" package will be used as a provided and fallback module
shareKey: 'shared-vue', // under this name the shared module will be placed in the share scope
shareScope: 'default', // share scope with this name will be used
singleton: true, // only a single version of the shared module is allowed
strictVersion: true, // don't use shared version when version isn't valid. Singleton or modules without fallback will throw, otherwise fallback is used
version: '1.2.3', // the version of the shared module
requiredVersion: '^1.0.0', // the required version of the shared module
},
},
}),
],
};
The NoEmitOnErrorsPlugin
allows you to avoid emitting assets when there are any errors. Enabled by default, you can disable using optimization.emitOnErrors
webpack.config.js
module.exports = {
plugins: [new webpack.NoEmitOnErrorsPlugin()],
};
The NormalModuleReplacementPlugin
allows you to replace resources that match resourceRegExp
with newResource
. If newResource
is relative, it is resolved relative to the previous resource. If newResource
is a function, it is expected to overwrite the request attribute of the supplied resource.
This can be useful for allowing different behaviour between builds.
new webpack.NormalModuleReplacementPlugin(resourceRegExp, newResource);
Replace a specific module when building for a development environment.
Say you have a configuration file some/path/config.development.module.js
and a special version for production in some/path/config.production.module.js
Add the following plugin when building for production:
new webpack.NormalModuleReplacementPlugin(
/some\/path\/config\.development\.js/,
'./config.production.js'
);
Conditional build depending on an specified environment.
Say you want a configuration with specific values for different build targets.
module.exports = function (env) {
var appTarget = env.APP_TARGET || 'VERSION_A';
return {
plugins: [
new webpack.NormalModuleReplacementPlugin(
/(.*)-APP_TARGET(\.*)/,
function (resource) {
resource.request = resource.request.replace(
/-APP_TARGET/,
`-${appTarget}`
);
}
),
],
};
};
Create the two configuration files:
app/config-VERSION_A.js
export default {
title: 'I am version A',
};
app/config-VERSION_B.js
export default {
title: 'I am version B',
};
Then import that configuration using the keyword you're looking for in the regexp:
import config from 'app/config-APP_TARGET';
console.log(config.title);
And now you get the right configuration imported depending on which target you're building for:
npx webpack --env APP_TARGET=VERSION_A
=> 'I am version A'
npx webpack --env APP_TARGET=VERSION_B
=> 'I am version B'
Prefetch normal module requests, causing them to be resolved and built before the first import
or require
of that module occurs. Using this plugin can boost performance. Try to profile the build first to determine clever prefetching points.
new webpack.PrefetchPlugin([context], request);
context
: An absolute path to a directoryrequest
: A request string for a normal moduleGenerate Chrome profile file which includes timings of plugins execution. Outputs events.json
file by default. It is possible to provide custom file path using outputPath
option.
Note : ProfilingPlugin accepts only absolute paths.
outputPath
: An absolute path to a custom output file (json)new webpack.debug.ProfilingPlugin();
outputPath
new webpack.debug.ProfilingPlugin({
outputPath: path.join(__dirname, 'profiling/profileEvents.json'),
});
In order to view the profile file:
ProfilingPlugin
.Performance
tab (formerly Timeline
).events.json
by default) into the profiler.It will then display timeline stats and calls per plugin!
object = { boolean activeModules = true, boolean entries = false, function (number percentage, string message, [string] ...args) handler, boolean modules = true, number modulesCount = 500, boolean profile = false }
function (number percentage, string message, [string] ...args)
The ProgressPlugin
provides a way to customize how progress is reported during a compilation.
Create an instance of ProgressPlugin
and provide one of the allowed params.
function
Provide a handler function which will be called when hooks report progress. handler
function arguments:
percentage
: a number between 0 and 1 indicating the completion percentage of the compilationmessage
: a short description of the currently-executing hook...args
: zero or more additional strings describing the current progressconst handler = (percentage, message, ...args) => {
// e.g. Output each progress message directly to the console:
console.info(percentage, message, ...args);
};
new webpack.ProgressPlugin(handler);
object
When providing an object
to the ProgressPlugin
, following properties are supported:
activeModules
(boolean = false
): Shows active modules count and one active module in progress message.entries
(boolean = true
): Shows entries count in progress message.handler
(See Providing function)modules
(boolean = true
): Shows modules count in progress message.modulesCount
(number = 5000
): A minimum modules count to start with. Takes effect when modules
property is enabled.profile
(boolean = false
): Tells ProgressPlugin
to collect profile data for progress steps.dependencies
(boolean = true
): Shows the count of dependencies in progress message.dependenciesCount
(number = 10000
): A minimum dependencies count to start with. Takes effect when dependencies
property is enabled.percentBy
(string = null: 'entries' | 'dependencies' | 'modules' | null
): Tells ProgressPlugin
how to calculate progress percentage.new webpack.ProgressPlugin({
activeModules: false,
entries: true,
handler(percentage, message, ...args) {
// custom logic
},
modules: true,
modulesCount: 5000,
profile: false,
dependencies: true,
dependenciesCount: 10000,
percentBy: null,
});
By default, progress percentage is calculated based on built modules count and total modules count: built / total
The total modules count is unknown in advance and changes during the build. This may cause inaccurate progress percentage.
To solve this problem ProgressPlugin
caches the last known total modules count and reuses this value on the next build. The first build will warm the cache but the following builds will use and update this value.
We recommend using
percentBy: 'entries'
setting for projects with multiple configured entry points. Percentage calculation will become more accurate because the amount of entry points is known in advance.
The following hooks report progress information to ProgressPlugin
.
Compiler
Compilation
Automatically load modules instead of having to import
or require
them everywhere.
new webpack.ProvidePlugin({
identifier: 'module1',
// ...
});
or
new webpack.ProvidePlugin({
identifier: ['module1', 'property1'],
// ...
});
By default, module resolution path is current folder (./**)
and node_modules
.
It is also possible to specify full path:
const path = require('path');
new webpack.ProvidePlugin({
identifier: path.resolve(path.join(__dirname, 'src/module1')),
// ...
});
Whenever the identifier
is encountered as free variable in a module, the module
is loaded automatically and the identifier
is filled with the exports of the loaded module
(or property
in order to support named exports).
For importing the default export of an ES2015 module, you have to specify the default property of module.
To automatically load jquery
we can point both variables it exposes to the corresponding node module:
new webpack.ProvidePlugin({
$: 'jquery',
jQuery: 'jquery',
});
Then in any of our source code:
// in a module
$('#item'); // <= works
jQuery('#item'); // <= also works
// $ is automatically set to the exports of module "jquery"
Angular looks for window.jQuery
in order to determine whether jQuery is present, see the source code.
new webpack.ProvidePlugin({
'window.jQuery': 'jquery',
});
new webpack.ProvidePlugin({
_map: ['lodash', 'map'],
});
new webpack.ProvidePlugin({
Vue: ['vue/dist/vue.esm.js', 'default'],
});
This plugin enables more fine grained control of source map generation. It is also enabled automatically by certain settings of the devtool
configuration option.
new webpack.SourceMapDevToolPlugin(options);
The following options are supported:
test
(string
RegExp
[string, RegExp]
): Include source maps for modules based on their extension (defaults to .js
, .mjs
, and .css
).include
(string
RegExp
[string, RegExp]
): Include source maps for module paths that match the given value.exclude
(string
RegExp
[string, RegExp]
): Exclude modules that match the given value from source map generation.filename
(string
): Defines the output filename of the SourceMap (will be inlined if no value is provided).append
(string
): Appends the given value to the original asset. Usually the #sourceMappingURL
comment. [url]
is replaced with a URL to the source map file. Since webpack v4.36.0, path parameters are supported: [chunk]
, [filename]
and [contenthash]
. Setting append
to false
disables the appending.moduleFilenameTemplate
(string
): See output.devtoolModuleFilenameTemplate
.fallbackModuleFilenameTemplate
(string
): See link above.namespace
(string
): See output.devtoolNamespace
.module = true
(boolean
): Indicates whether loaders should generate source maps.columns = true
(boolean
): Indicates whether column mappings should be used.noSources = false
(boolean
): Prevents the source file content from being included in the source map.publicPath
(string
): Emits absolute URLs with public path prefix, e.g. https://example.com/project/
.fileContext
(string
): Makes the [file]
argument relative to this directory.sourceRoot
(string
): Provide a custom value for the sourceRoot
property in the SourceMap.The fileContext
option is useful when you want to store source maps in an upper level directory to avoid ../../
appearing in the absolute [url]
.
The following examples demonstrate some common use cases for this plugin.
You can use the following code to replace the configuration option devtool: inline-source-map
with an equivalent custom plugin configuration:
module.exports = {
// ...
devtool: false,
plugins: [new webpack.SourceMapDevToolPlugin({})],
};
The following code would exclude source maps for any modules in the vendor.js
bundle:
new webpack.SourceMapDevToolPlugin({
filename: '[name].js.map',
exclude: ['vendor.js'],
});
Set a URL for source maps. Useful for hosting them on a host that requires authorization.
new webpack.SourceMapDevToolPlugin({
append: '\n//# sourceMappingURL=https://example.com/sourcemap/[url]',
filename: '[name].map',
});
And for cases when source maps are stored in the upper level directory:
project
|- dist
|- public
|- bundle-[hash].js
|- sourcemaps
|- bundle-[hash].js.map
With next config:
new webpack.SourceMapDevToolPlugin({
filename: 'sourcemaps/[file].map',
publicPath: 'https://example.com/project/',
fileContext: 'public',
});
Will produce the following URL:
https://example.com/project/sourcemaps/bundle-[hash].js.map
Originally, chunks (and modules imported inside them) were connected by a parent-child relationship in the internal webpack graph. The CommonsChunkPlugin
was used to avoid duplicated dependencies across them, but further optimizations were not possible.
Since webpack v4, the CommonsChunkPlugin
was removed in favor of optimization.splitChunks
.
Out of the box SplitChunksPlugin
should work well for most users.
By default it only affects on-demand chunks, because changing initial chunks would affect the script tags the HTML file should include to run the project.
webpack will automatically split chunks based on these conditions:
node_modules
folderWhen trying to fulfill the last two conditions, bigger chunks are preferred.
webpack provides a set of options for developers that want more control over this functionality.
optimization.splitChunks
This configuration object represents the default behavior of the SplitChunksPlugin
.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
chunks: 'async',
minSize: 20000,
minRemainingSize: 0,
minChunks: 1,
maxAsyncRequests: 30,
maxInitialRequests: 30,
enforceSizeThreshold: 50000,
cacheGroups: {
defaultVendors: {
test: /[\\/]node_modules[\\/]/,
priority: -10,
reuseExistingChunk: true,
},
default: {
minChunks: 2,
priority: -20,
reuseExistingChunk: true,
},
},
},
},
};
splitChunks.automaticNameDelimiter
string = '~'
By default webpack will generate names using origin and name of the chunk (e.g. vendors~main.js
). This option lets you specify the delimiter to use for the generated names.
splitChunks.chunks
string = 'async'
function (chunk)
This indicates which chunks will be selected for optimization. When a string is provided, valid values are all
, async
, and initial
. Providing all
can be particularly powerful, because it means that chunks can be shared even between async and non-async chunks.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
// include all types of chunks
chunks: 'all',
},
},
};
Alternatively, you may provide a function for more control. The return value will indicate whether to include each chunk.
module.exports = {
//...
optimization: {
splitChunks: {
chunks(chunk) {
// exclude `my-excluded-chunk`
return chunk.name !== 'my-excluded-chunk';
},
},
},
};
splitChunks.maxAsyncRequests
number = 30
Maximum number of parallel requests when on-demand loading.
splitChunks.maxInitialRequests
number = 30
Maximum number of parallel requests at an entry point.
splitChunks.defaultSizeTypes
[string] = ['javascript', 'unknown']
Sets the size types which are used when a number is used for sizes.
splitChunks.minChunks
number = 1
The minimum times must a module be shared among chunks before splitting.
splitChunks.hidePathInfo
boolean
Prevents exposing path info when creating names for parts splitted by maxSize.
splitChunks.minSize
number = 20000
Minimum size, in bytes, for a chunk to be generated.
splitChunks.enforceSizeThreshold
splitChunks.cacheGroups.{cacheGroup}.enforceSizeThreshold
number = 50000
Size threshold at which splitting is enforced and other restrictions (minRemainingSize, maxAsyncRequests, maxInitialRequests) are ignored.
splitChunks.minRemainingSize
splitChunks.cacheGroups.{cacheGroup}.minRemainingSize
number = 0
splitChunks.minRemainingSize
option was introduced in webpack 5 to avoid zero sized modules by ensuring that the minimum size of the chunk which remains after splitting is above a limit. Defaults to 0
in 'development' mode. For other cases splitChunks.minRemainingSize
defaults to the value of splitChunks.minSize
so it doesn't need to be specified manually except for the rare cases where deep control is required.
splitChunks.layer
splitChunks.cacheGroups.{cacheGroup}.layer
RegExp
string
function
Assign modules to a cache group by module layer.
splitChunks.maxSize
number = 0
Using maxSize
(either globally optimization.splitChunks.maxSize
per cache group optimization.splitChunks.cacheGroups[x].maxSize
or for the fallback cache group optimization.splitChunks.fallbackCacheGroup.maxSize
) tells webpack to try to split chunks bigger than maxSize
bytes into smaller parts. Parts will be at least minSize
(next to maxSize
) in size.
The algorithm is deterministic and changes to the modules will only have local effects. So that it is usable when using long term caching and doesn't require records. maxSize
is only a hint and could be violated when modules are bigger than maxSize
or splitting would violate minSize
.
When the chunk has a name already, each part will get a new name derived from that name. Depending on the value of optimization.splitChunks.hidePathInfo
it will add a key derived from the first module name or a hash of it.
maxSize
option is intended to be used with HTTP/2 and long term caching. It increases the request count for better caching. It could also be used to decrease the file size for faster rebuilding.
splitChunks.maxAsyncSize
number
Like maxSize
, maxAsyncSize
can be applied globally (splitChunks.maxAsyncSize
), to cacheGroups (splitChunks.cacheGroups.{cacheGroup}.maxAsyncSize
), or to the fallback cache group (splitChunks.fallbackCacheGroup.maxAsyncSize
).
The difference between maxAsyncSize
and maxSize
is that maxAsyncSize
will only affect on-demand loading chunks.
splitChunks.maxInitialSize
number
Like maxSize
, maxInitialSize
can be applied globally (splitChunks.maxInitialSize
), to cacheGroups (splitChunks.cacheGroups.{cacheGroup}.maxInitialSize
), or to the fallback cache group (splitChunks.fallbackCacheGroup.maxInitialSize
).
The difference between maxInitialSize
and maxSize
is that maxInitialSize
will only affect initial load chunks.
splitChunks.name
boolean = false
function (module, chunks, cacheGroupKey) => string
string
Also available for each cacheGroup: splitChunks.cacheGroups.{cacheGroup}.name
.
The name of the split chunk. Providing false
will keep the same name of the chunks so it doesn't change names unnecessarily. It is the recommended value for production builds.
Providing a string or a function allows you to use a custom name. Specifying either a string or a function that always returns the same string will merge all common modules and vendors into a single chunk. This might lead to bigger initial downloads and slow down page loads.
If you choose to specify a function, you may find the chunk.name
and chunk.hash
properties (where chunk
is an element of the chunks
array) particularly useful in choosing a name for your chunk.
If the splitChunks.name
matches an entry point name, the entry point will be removed.
main.js
import _ from 'lodash';
console.log(_.join(['Hello', 'webpack'], ' '));
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
commons: {
test: /[\\/]node_modules[\\/]/,
// cacheGroupKey here is `commons` as the key of the cacheGroup
name(module, chunks, cacheGroupKey) {
const moduleFileName = module
.identifier()
.split('/')
.reduceRight((item) => item);
const allChunksNames = chunks.map((item) => item.name).join('~');
return `${cacheGroupKey}-${allChunksNames}-${moduleFileName}`;
},
chunks: 'all',
},
},
},
},
};
Running webpack with following splitChunks
configuration would also output a chunk of the group common with next name: commons-main-lodash.js.e7519d2bb8777058fa27.js
(hash given as an example of real world output).
splitChunks.usedExports
splitChunks.cacheGroups{cacheGroup}.usedExports
boolean = true
Figure out which exports are used by modules to mangle export names, omit unused exports and generate more efficient code.
When it is true
: analyse used exports for each runtime, when it is "global"
: analyse exports globally for all runtimes combined).
splitChunks.cacheGroups
Cache groups can inherit and/or override any options from splitChunks.*
; but test
, priority
and reuseExistingChunk
can only be configured on cache group level. To disable any of the default cache groups, set them to false
.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
default: false,
},
},
},
};
splitChunks.cacheGroups.{cacheGroup}.priority
number = -20
A module can belong to multiple cache groups. The optimization will prefer the cache group with a higher priority
. The default groups have a negative priority to allow custom groups to take higher priority (default value is 0
for custom groups).
splitChunks.cacheGroups.{cacheGroup}.reuseExistingChunk
boolean = true
If the current chunk contains modules already split out from the main bundle, it will be reused instead of a new one being generated. This can affect the resulting file name of the chunk.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
reuseExistingChunk: true,
},
},
},
},
};
splitChunks.cacheGroups.{cacheGroup}.type
function
RegExp
string
Allows to assign modules to a cache group by module type.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
json: {
type: 'json',
},
},
},
},
};
splitChunks.cacheGroups.test
splitChunks.cacheGroups.{cacheGroup}.test
function (module, { chunkGraph, moduleGraph }) => boolean
RegExp
string
Controls which modules are selected by this cache group. Omitting it selects all modules. It can match the absolute module resource path or chunk names. When a chunk name is matched, all modules in the chunk are selected.
Providing a function to{cacheGroup}.test
:
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
svgGroup: {
test(module) {
// `module.resource` contains the absolute path of the file on disk.
// Note the usage of `path.sep` instead of / or \, for cross-platform compatibility.
const path = require('path');
return (
module.resource &&
module.resource.endsWith('.svg') &&
module.resource.includes(`${path.sep}cacheable_svgs${path.sep}`)
);
},
},
byModuleTypeGroup: {
test(module) {
return module.type === 'javascript/auto';
},
},
},
},
},
};
In order to see what information is available in module
and chunks
objects, you can put debugger;
statement in the callback. Then run your webpack build in debug mode to inspect the parameters in Chromium DevTools.
Providing a RegExp
to {cacheGroup}.test
:
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
// Note the usage of `[\\/]` as a path separator for cross-platform compatibility.
test: /[\\/]node_modules[\\/]|vendor[\\/]analytics_provider|vendor[\\/]other_lib/,
},
},
},
},
};
splitChunks.cacheGroups.{cacheGroup}.filename
string
function (pathData, assetInfo) => string
Allows to override the filename when and only when it's an initial chunk.
All placeholders available in output.filename
are also available here.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
filename: '[name].bundle.js',
},
},
},
},
};
And as a function:
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
filename: (pathData) => {
// Use pathData object for generating filename string based on your requirements
return `${pathData.chunk.name}-bundle.js`;
},
},
},
},
},
};
It is possible to create a folder structure by providing path prefixing the filename: 'js/vendor/bundle.js'
.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
filename: 'js/[name]/bundle.js',
},
},
},
},
};
splitChunks.cacheGroups.{cacheGroup}.enforce
boolean = false
Tells webpack to ignore splitChunks.minSize
, splitChunks.minChunks
, splitChunks.maxAsyncRequests
and splitChunks.maxInitialRequests
options and always create chunks for this cache group.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
enforce: true,
},
},
},
},
};
splitChunks.cacheGroups.{cacheGroup}.idHint
string
Sets the hint for chunk id. It will be added to chunk's filename.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
defaultVendors: {
idHint: 'vendors',
},
},
},
},
};
// index.js
import('./a'); // dynamic import
// a.js
import 'react';
//...
Result: A separate chunk would be created containing react
. At the import call this chunk is loaded in parallel to the original chunk containing ./a
.
Why:
node_modules
react
is bigger than 30kbWhat's the reasoning behind this? react
probably won't change as often as your application code. By moving it into a separate chunk this chunk can be cached separately from your app code (assuming you are using chunkhash, records, Cache-Control or other long term cache approach).
// entry.js
// dynamic imports
import('./a');
import('./b');
// a.js
import './helpers'; // helpers is 40kb in size
//...
// b.js
import './helpers';
import './more-helpers'; // more-helpers is also 40kb in size
//...
Result: A separate chunk would be created containing ./helpers
and all dependencies of it. At the import calls this chunk is loaded in parallel to the original chunks.
Why:
helpers
is bigger than 30kbPutting the content of helpers
into each chunk will result into its code being downloaded twice. By using a separate chunk this will only happen once. We pay the cost of an additional request, which could be considered a tradeoff. That's why there is a minimum size of 30kb.
Create a commons
chunk, which includes all code shared between entry points.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
commons: {
name: 'commons',
chunks: 'initial',
minChunks: 2,
},
},
},
},
};
Create a vendors
chunk, which includes all code from node_modules
in the whole application.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
commons: {
test: /[\\/]node_modules[\\/]/,
name: 'vendors',
chunks: 'all',
},
},
},
},
};
Create a custom vendor
chunk, which contains certain node_modules
packages matched by RegExp
.
webpack.config.js
module.exports = {
//...
optimization: {
splitChunks: {
cacheGroups: {
vendor: {
test: /[\\/]node_modules[\\/](react|react-dom)[\\/]/,
name: 'vendor',
chunks: 'all',
},
},
},
},
};
Ignore the specified files, i.e. those matching the provided paths or regular expressions, while in watch mode.
new webpack.WatchIgnorePlugin({ paths });
paths
(Array<string | RegExp>
): A list of RegExps or absolute paths to directories or files that should be ignored.