browserize

browserize

Converts simple node.js modules into ES6 modules.

• What it is
  • What it does
  • What it does not
  • When to use
  • When not to use
  • "But browserize does almost what I need"
• How to use it
  • CLI
  • node API
  • Requirements

---

What it is

• What it does
• What it does not
• When to use
• When not to use
• "But browserize does almost what I need"

What it does

browserize turns this:

module.exports = function defaultExport() {}

into this:

export default function defaultExport() {}

What it does not

browserize does not:

• check if the result will run in a browser
• transform requires into imports
• bundle dependencies àla Webpack/Rollup
• transpile anything other than JavaScript, like CoffeeScript (it might work by coincidence, but there's no support for that)

When to use

browserize is made for small packages without dependencies that should run both in node.js and in the browser.

When not to use

If your package has any dependency, it's probably complex enough to warrant babel, webpack, or some such. Use that instead.

If you need to transpile anything, like CoffeScript or TypeScript, your tooling for that should cover you.

"But browserize does almost what I need"

Open an issue, and let's talk about it 😉

How to use it

• node API
• CLI
• Requirements

node API

browserize takes an options object with three optional entries:

• default: the file where the default export is found, defaults to index.js
• named: where to find the named exports, defaults to null
• output: where to write the ESM file, defaults to the default or named filename with the extension .mjs

And that is it.

Two examples

The simplest form

const browserize = require('browserize')
browserize()

This reads index.js and writes the equivalent index.mjs, and that's it.

The most complex case browserize covers

const browserize = require('browserize')

browserize({
	default: 'class-name.jsx',
	named: 'helper-functions.js',
	output: 'dist/browser-magic.js',
})

This includes named exports and sets custom paths for everything.

CLI

npx browserize [--no-default|-x] [[--default|-d] index.js] [[--named|-n] helpers.js] [[--output|-o] index.mjs]

The CLI passes the given arguments through to the underlying node API.

Three examples

The simplest form

npx browserize

This reads index.js and writes the equivalent index.mjs, and that's it.

Adding named exports

npm browserize -n helper-functions

This reads index.js and helper-functions.js, then transforms concatenates them, and finally writes the result to index.mjs.

The most complex case browserize covers

npx browserize class-name.jsx helper-functions.js dist/browser-magic.js

This includes named exports and sets custom paths for everything.

Requirements

browserize is a simple tool and has a few simple requirements:

Each source file must contain exactly one assignment to module.exports

Good

module.exports = class DefaultExport {}

module.exports = {
	key1: helper1,
	key2: helper2,
}

Bad

exports.key1 = helper1
exports.key2 = helper2

While valid, browserize does not know how to transform this.

module.exports = export1
module.exports = export2

This is not useful anyway.

window.myStuff = class DefaultExport {}

This is not a module.

The default export must be declared without a newline between the assignment operator and the exported item

Good

module.exports = class DefaultExport {}

module.exports = class DefaultExport {
}

Bad

module.exports =
class DefaultExport {}

While this is valid in node.js, it will lead to an invalid ESM file.

The named exports must be declared as an object literal

Good

module.exports = { helper1, helper2 }

module.exports = {
	key1: helper1,
	key2: helper2,
}

Bad

module.exports.key1 = helper1
module.exports.key2 = helper2

While this is valid in node.js, browserize does not understand it.

This is too complex, and has no real benefit over the object literal.