babel-plugin-i18next-extract

babel-plugin-i18next-extract

---

babel-plugin-i18next-extract is a Babel Plugin that will traverse your Javascript/Typescript code in order to find i18next translation keys.

Features

• ☑️ Keys extraction in JSON v3 format.
• ☑️ Detection of i18next.t() function calls.
• ☑️ Full react-i18next support.
• ☑️ Plurals support.
• ☑️ Contexts support.
• ☑️ Namespace detection.
• ☑️ Disable extraction on a specific file sections or lines using comment hints.
• ☑️ Overwrite namespaces, plurals and contexts on-the-fly using comment hints.
• … and more?

Installation

yarn add --dev babel-plugin-i18next-extract

# or

npm i --save-dev babel-plugin-i18next-extract

Usage

If you already use Babel, chances are you already have a babel configuration (e.g. a .babelrc file). Just declare the plugin and you're good to go:

{
  "plugins": [
    "i18next-extract",
    // your other plugins…
  ]
}

You may want to specify additional configuration options:

{
  "plugins": [
    ["i18next-extract", {"nsSeparator": "~"}],
    // your other plugins…
  ]
}

Once the plugin is setup, you can build your app normally or run Babel through Babel CLI:

yarn run babel -f .babelrc 'src/**/*.{js,jsx,ts,tsx}'

# or

npm run babel -f .babelrc 'src/**/*.{js,jsx,ts,tsx}'

Extracted translations should land in the extractedTranslations/ directory by default.

If you don't have a babel configuration yet, you can follow the Configure Babel documentation to get started.

Configuration

Option	Type	Description	Default
locales	string[]	Locales your project supports.	['en']
defaultNS	string	The default namespace that your translation use.	'translation'
pluralSeparator	string	String you want to use to split plural from keys. See i18next Configuration options	'_'
contextSeparator	string	String you want to use to split context from keys. See i18next Configuration options	'_'
keySeparator	string or null	String you want to use to split keys. Set to null if you don't want to split your keys or if you want to use keys as value. See i18next Configuration options	'.'
nsSeparator	string or null	String you want to use to split namespace from keys. Set to null if you don't want to infer a namespace from key value or if you want to use keys as value. See See i18next Configuration options	':'
i18nextInstancesNames	string[]	Possible names for your i18next instances. This will be used to detect i18next.t calls.	['i18next', 'i18n']
tFunctionNames	string[]	Possible names for your t functions. This will only be used for direct calls to t functions (i.e. t('key'), not foo.t('key')) and in very last resort.	['t']
defaultContexts	string[]	Default context keys to create when detecting a translation with context.	['', '_male', '_female']
outputPath	string	Path where translation keys should be extracted to. You can put {{ns}} and {{locale}} placeholders in the value to change the location depending on the namespace or the locale.	extractedTranslations/{{locale}}/{{ns}}.json
defaultValue	string or null	Default value for extracted keys.	'' (empty string)
keyAsDefaultValue	boolean or string[]	If true, use the extracted key as defaultValue (ignoring defaultValue option). You can also specify an array of locales to apply this behavior only to a specific set locales (e.g. if you keys are in plain english, you may want to set this option to ['en']).	false
keyAsDefaultValueForDerivedKeys	boolean	If false and keyAsDefaultValue is enabled, don't use derived keys (plural forms or contexts) as default value. defaultValue will be used instead.	true
discardOldKeys	boolean	When set to true, keys that no longer exist are removed from the JSON files. By default, new keys will be added to the JSON files and never removed.	false
jsonSpace	number	Number of indentation space to use in extracted JSON files.	2

Comment hints

Disable extraction on a specific section

If the plugin extracts a key you want to skip or erroneously tries to parse a function that doesn't belong to i18next, you can use a comment hint to disable the extraction:

// i18next-extract-disable-next-line
i18next.t("this key won't be extracted")

i18next.t("neither this one") // i18next-extract-disable-line

// i18next-extract-disable
i18next.t("or this one")
i18next.t("and this one")
// i18next-extract-enable

i18next.t("but this one will be")

You can put a // i18next-extract-disable comment at the top of the file in order to disable extraction on the entire file.

Explicitly specify contexts for a key

This is very useful if you want to use different contexts than the default male and female for a given key:

// i18next-extract-mark-context-next-line ["dog", "cat"]
i18next.t("this key will have dog and cat context", {context: dogOrCat})

// i18next-extract-mark-context-next-line
i18next.t("this key will have default context, although no context is specified")

// i18next-extract-mark-context-next-line disable
i18next.t("this key wont have a context, although a context is specified", {context})

i18next.t("can be used on line") // i18next-extract-mark-context-line

// i18next-extract-mark-context-start
i18next.t("or on sections") 
// i18next-extract-mark-context-stop

const transComponent = (
  // i18next-extract-mark-context-next-line
  <Trans>it also works on Trans components</Trans>
)

Explicitly use a namespace for a key

// i18next-extract-mark-ns-next-line forced-ns
i18next.t("this key will be in forced-ns namespace")

i18next.t("this one also", {ns: 'this-ns-wont-be-used'}) // i18next-extract-mark-ns-line forced-ns

// i18next-extract-mark-ns-start forced-ns
i18next.t("and still this one")
// i18next-extract-mark-ns-stop forced-ns

Explicitly enable/disable a plural form for a key

// i18next-extract-mark-plural-next-line
i18next.t("this key will be in forced in plural form")

// i18next-extract-mark-plural-next-line disable
i18next.t("this key wont have plural form", {count})

Usage with create-react-app

create-react-app doesn't let you modify the babel configuration. Fortunately, it's still possible to use this plugin without ejecting. First of all, install Babel CLI:

yarn add --dev @babel/cli

# or

npm add --save-dev @babel/cli

Create a minimal .babelrc that uses the react-app babel preset (DO NOT install it, it's already shipped with CRA):

{
  "presets": ["react-app"],
  "plugins": ["i18next-extract"]
}

You should then be able to extract your translations using the CLI:

# NODE_ENV must be specified for react-app preset to work properly
NODE_ENV=development yarn run babel -f .babelrc 'src/**/*.{js,jsx,ts,tsx}'

To simplify the extraction, you can add a script to your package.json:

"scripts": {
  […]
  "i18n-extract": "NODE_ENV=development babel -f .babelrc 'src/**/*.{js,jsx,ts,tsx}'",
  […]
}

And then just run:

yarn run i18n-extract

# or

npm run i18n-extract

Gotchas

The plugin tries to be a little smart, but can't do magic. i18next has a runtime unlike this plugin which must guess everything statically. For instance, you may want to disable extraction on dynamic keys:

i18next.t(myVariable);
i18next.t(`error.${code}`);

If you try to extract keys from this code, the plugin will issue a warning because it won't be able to infer the translations statically. If you really want to specify variable keys, you should skip them with a comment hint. The same goes for plural forms, context and namespace detection:

i18next.t("myKey", myOpts); // This won't work.

However, in React components, it might come handy to be a little smarter than that. That's why when using a <Trans> component, the plugin will try to resolve references before resigning. For instance, this code should extract properly:

const foo = <p>Hello</p>
const bar = <Trans>{foo} world</Trans>