An open API service indexing awesome lists of open source software.

https://github.com/offen/l10nify

gettext based localization workflow for Javascript / Browserify
https://github.com/offen/l10nify

browserify browserify-transform gettext i18n javascript localization xgettext

Last synced: 7 months ago
JSON representation

gettext based localization workflow for Javascript / Browserify

Awesome Lists containing this project

README

          


offen.software logo

# l10nify

Localization workflow for Browserify

## How does this work?

This is a Browserify-based workflow for localizing client side applications. It is built for use cases where you can and want to ship one bundle per locale. Strings are defined by calling a pseudo-global function (i.e. `__(string, args..)`) in the default language in your code (similar to GNU gettext or similar). Strings are stored in `.po` files.

## Installation

Install from npm:

```
npm install @offen/l10nify
```

This installs two things:

- a Browserify transform for localizing strings at bundle time. It is references as `@offen/l10nify`
- a `extract-strings` command that you can use to generate PO files from your JavaScript code.

## Usage

### Defining strings in client side code

In your code, use the `__(string, args...)` function (`__` is the default, but you can use anything) to declare strings in your default language (which defaults to `en` but can be anything you want it to):

```js
const HappyBirthdayComponent = (props) => {
return (

{__('Happy birthday, %s!', props.name)}


)
}
```

### Extract strings from your code

Next, you can extract these strings from your code into `.po` files using the `extract-strings` command:

```
$(npm bin)/extract-strings **/*.js
```

This will extract the strings from all matching files and print a `.po` file to stdout. Use the standard `gettext` tools like `msgmerge` and `msgcat` to combine the output with existing `.po` files.

Refer to `extract-strings --help` for a full list of options

#### Why not just use `xgettext`

While `xgettext` works perfectly fine on ES5 code, it will choke on ES6+ syntax and also does not support parsing JSX, which `l10nify` supports.

### Apply the transform at bundle time

Apply the transform to your Browserify setup passing the target `locale`. In development, you can omit this parameter to make the transform return the default locale, i.e. the strings you defined in code.

```js
var browserify = require('browserify')

var b = browserify()
b.add('app.js')
b.transform('@offen/l10nify', { locale: 'fr' })
b.bundle(function (err, src) {
// consume bundle
})
```

#### Options

The following options can be passed to the transform:

##### `locale`

`locale` specifies the locale which you want to return when bundling. It defaults to `en` or `process.env.LOCALE` when set.

##### `defaultLocale`

`defaultLocale` specifies the default locale that is used to define strings in code. It defaults to `en` or `process.env.DEFAULT_LOCALE` when set.

##### `source`

`source` specifies the directory in which the `.po` files are stored. It defaults to `./locales` or `process.env.SOURCE` when set.

##### `global`

`global` defines the global function identifier that is used for defining strings in code. It defaults to `__` or `process.env.GLOBAL` when set.

## License

Copyright 2020 Frederik Ring - `l10nify` is available under the MIT License