Multiple entry points can be useful if your application has multiple ways it can be entered but share a common source tree. Such as if creating an optimized and separate build for mobile devices or builds specific for users of a certain language.
Multiple entry points are specified by supplying an array or object to the entry
configuration option in your webpack.config.js
:
module.exports = {
entry: {
desktop: './src/desktop.js',
mobile: './src/mobile.js'
},
output: {
path: './dist',
filename: '[name].bundle.js',
chunkFilename: '[id].common.js'
}
};
This will create two separate builds within the ./dist
folder named: ./dist/desktop.bundle.js
and ./dist/mobile.bundle.js
respectively. Any common parts of the application that are loaded asynchronously will be bundled to ./dist/1.common.js
(with the number increasing as the number of chunks increases). Each of the entry points of the build will share this code and only load each chunk as needed.
webpack includes another directive '[hash]'
to generate a hash name based upon your build. This is useful in conjunction with setting a long expiration time as the file is served, such as 1 year. Consumers of your app will hopefully cache your application and quickly reload from their local copy upon subsequent visits. Then when your build has changed and a new hash is generated, consumers be given the new file and the latest version of your application.
module.exports = {
entry: './src/main.js',
output: {
path: './dist',
filename: '[hash].js'
}
};
Will generate the a file named similar to: ./dist/47a0cc1b840cb310842cb85fb5b6116c.js
upon bundling.
// TODO: Talk about the recordsPath
configuration option.