This is an i18next backend plugin to be used for locize service. It will load resources from locize server using http fetch or xhr as fallback.
If you're not familiar with i18next and how i18next backend plugins works, please first have a look at the i18next documentation.
It will allow you to save missing keys containing both default value and context information by calling:
i18next.t(key, defaultValue, tDescription);
i18next.t(key, { defaultValue, tDescription });
To see i18next-locize-backend in a working app example, you may have a look at:
Make sure you set the debug option of i18next to true. This will maybe log more information in the developer console.
SaveMissing is not working
Did you wait 5-10 seconds before refreshing the locize UI? It may take a couple of seconds until the missing keys are sent and saved.
Per default only localhost is allowed to send missing keys (or update missing keys) (to avoid using this feature accidentally in production). If you're not using localhost during development you will have to set the allowedAddOrUpdateHosts: ['your.domain.tld'] for the backend options.
It's also recommended to set the fallbackLng equal to the source language defined in locize. i.e. if your source language in locize is de, set the fallbackLng also to de.
Loading translations not working
Make sure the translations are published, either by having enabled auto publishing for your version or by manually publishing the version. Alternatively, you can publish via CLI or directly by consuming the API.
In case you're using the private publish mode, make sure you're using the correct api key and are setting the private option to true.
import i18next from "i18next";
import Locize from "i18next-locize-backend";
i18next.use(Locize).init({
backend: {
projectId: "[PROJECTID]",
apiKey: "[APIKEY]",
version: "[VERSION]",
private: true,
referenceLng: "en",
cdnType: "standard"
},
fallbackLng: 'en', // typically your reference language in locize
supportedLngs: ['en', 'de'] // should match your available languages in locize
});
On server side: process is not exiting
In case you want to use i18next-locize-backend on server side for a short running process, you might want to set the reloadInterval option to false:
{
reloadInterval: false,
projectId: "[PROJECTID]",
version: 'latest',
referenceLng: 'en',
cdnType: 'standard'
}
On server side: downloads keep growing
On the server reloadInterval defaults to 1 hour, and the timer that drives it keeps the i18next instance alive for as long as the process runs. That is what you want for a long-lived singleton, and it is a trap for anything short-lived: if you create a new i18next instance per request or per render (a per-request instance, an SSR pass of a component that initializes i18next, a job that runs on a schedule), every one of them leaves a live timer behind and keeps refetching your namespaces forever. Nothing errors and the process still exits normally, because the timer is unrefed. The only symptom is your download count climbing between restarts.
So: create the instance once per process and reuse it, or set reloadInterval: false on instances that are not the singleton.
{
reloadInterval: false, // this instance is not a long-lived singleton
projectId: "[PROJECTID]",
version: 'latest'
}
If you are unsure whether this is happening to you, check the Downloads by Browser chart on your locize project's metrics page: server-side requests show up as i18next-locize-backend, and a per-instance leak looks like a floor that ratchets up between deploys instead of following your traffic.
Not all languages are loaded
By default the supportedLngs are defined by having a minimum of 90% of done translations. You can set a threshold for languages to be added to supportedLngs by setting translatedPercentageThreshold in backend options (eg: 1 = 100% translated, 0.9 = 90% translated).
Source can be loaded via npm, yarn, or downloaded from this repo.
# npm package
$ npm install i18next-locize-backend
# yarn
$ yarn add i18next-locize-backend
v10 requires native
fetch. Node ≥ 18, all modern browsers, Deno, and Bun shipfetchby default — no extra setup needed. v10 dropped the bundledcross-fetchfallback that v9 used. For runtimes without nativefetch, install a ponyfill yourself before loading this backend, or stay on v9.
Wiring up:
import i18next from 'i18next';
import Locize from 'i18next-locize-backend';
// or
const i18next = require('i18next');
const Locize = require('i18next-locize-backend');
i18next.use(Locize).init(i18nextOptions);
for Deno:
import i18next from 'https://deno.land/x/i18next/index.js'
import Backend from 'https://deno.land/x/i18next_locize_backend/index.js'
i18next.use(Backend).init(i18nextOptions);
window.i18nextLocizeBackendnew Locize(options, callback) constructor or the getOptions function (see below) in combination with a language detector, you should define the supportedLngs option yourself in the i18next init options. Otherwise, i18next might try to load languages that do not exist in your project, resulting in unnecessary requests.IMPORTANT make sure you do not add your apiKey in the production build to avoid misuse by strangers
{
// the id of your locize project
projectId: '[PROJECTID]',
// add an api key if you want to send missing keys
apiKey: '[APIKEY]',
// the reference language of your project
referenceLng: '[LNG]',
// version - defaults to latest
version: '[VERSION]',
// configures the cdn endpoint that should be used (depends on which cdn type you've in your locize project) - defaults to 'standard'
cdnType: 'pro',
// private - set to true if you version on locize is set to use private publish
private: false,
// hostnames that are allowed to create, update keys
// please keep those to your local system, staging, test servers (not production)
// can be array of allowed hosts or a function (hostname) => { return true; // or false if not allowed }
allowedAddOrUpdateHosts: ['localhost'],
// optional event triggered on saved to backend
onSaved: (lng, ns) => { ... },
// can be used to reload resources in a specific interval (useful in server environments)
reloadInterval: typeof window !== 'undefined' ? false : 60 * 60 * 1000,
// define the threshold for languages to be added to supportedLngs (eg: 1 = 100% translated, 0.9 = 90% translated [default]).
translatedPercentageThreshold: 0.8,
// define a custom request function
// can be used to support Angular http client
//
// 'info' contains 'url', 'method', 'body' and 'headers'
// 'url' the url that should be requested
// 'method' GET for fetching translations and POST for saving missing translations
// 'body' will be a key:value object used when saving missing translations
// 'headers' will be a key:value object containing the header information that should be sent
// 'callback' is a function that takes two parameters, 'err' and 'res'.
// 'err' should be an error
// 'res' should be an object with a 'status' property and a 'data' property containing a stringified object instance beeing the key:value translation pairs for the
// requested language and namespace, or null in case of an error.
request: function (info, callback) {},
// or async / promise
//request: async (info) {},
/**
* During development this option can be set to true, to get uncached translations.
* If not set, it will be default true, if the i18next debug option is true, else default false.
* Do NOT set this to true on production!
*/
noCache: false,
/**
* Enable a small in-memory Cache-Control-aware cache for runtimes that
* don't have a native HTTP cache (Node.js, Deno, React Native, etc).
*
* - Only affects GET requests (POST never cached).
* - Parses "Cache-Control: max-age" and stores the full response until expiry.
* - Default: true for non-browser environments, false for browsers.
*
* Note: cache is process-local and unbounded — disable for long-running apps
* that fetch many dynamic URLs.
*/
useCacheLayer: true,
}
To load translations only projectId needs to be filled. To use the saveMissing feature of i18next additional to the projectId both apiKey and referenceLng have to be set.
Options can be passed in:
preferred - by setting options.backend in i18next.init:
import i18next from "i18next";
import Locize from "i18next-locize-backend";
i18next.use(Locize).init({
backend: options,
fallbackLng: 'en', // typically your reference language in locize
supportedLngs: ['en', 'de'] // should match your available languages in locize
});
on construction:
import Locize from "i18next-locize-backend";
const locize = new Locize(options);
via calling init:
import Locize from "i18next-locize-backend";
const locize = new Locize();
locize.init(options);
Will return a list of all languages in your project including percentage of translations done per version.
import Locize from "i18next-locize-backend";
const locize = new Locize(options);
locize.getLanguages((err, data) => {
/*
data is:
{
"en": {
"name": "English",
"nativeName": "English",
"isReferenceLanguage": true,
"translated": {
"latest": 1
}
},
"de": {
"name": "German",
"nativeName": "Deutsch",
"isReferenceLanguage": false,
"translated": {
"latest": 0.9
}
}
}
*/
});
// or
const data = await locize.getLanguages();
// or
i18next.services.backendConnector.backend.getLanguages(callback);
// or
const data = await i18next.services.backendConnector.backend.getLanguages();
Will return an object containing useful informations for the i18next init options.
import Locize from "i18next-locize-backend";
const locize = new Locize(options);
locize.getOptions((err, data) => {
/*
data is:
{
fallbackLng: 'en',
referenceLng: 'en',
supportedLngs: ['en', 'de'],
load: 'languageOnly|all' // depending on your supportedLngs has locals having region like en-US
}
*/
});
// or
const data = await locize.getOptions();
// or
i18next.services.backendConnector.backend.getOptions(callback);
// or
const data = await i18next.services.backendConnector.backend.getOptions();
You can set a threshold for languages to be added to supportedLngs by setting translatedPercentageThreshold in backend options (eg: 1 = 100% translated, 0.9 = 90% translated).
You can load some information from the backend to eg. set supportedLngs for i18next just supporting languages you got in your locize project.
You will get i18next options for (same as above backend.getOptions):
import i18next from "i18next";
import Locize from "i18next-locize-backend";
const locize = new Locize(
{
projectId: "[PROJECTID]",
apiKey: "[APIKEY]",
version: "[VERSION]"
// referenceLng -> not needed as will be loaded from API
},
(err, opts, lngs) => {
i18next.use(locize).init({ ...opts, ...yourOptions }); // yourOptions should not include backendOptions!
}
);
Use setI18n to pass in the i18next instance before initializing:
import i18n from "i18next";
import { initReactI18next, setI18n } from "react-i18next";
import LocizeBackend from "i18next-locize-backend";
const backendOptions = {
projectId: "1d0aa5aa-4660-4154-b6d9-907dbef10bb3"
};
const yourOptions = {
debug: true,
fallbackLng: 'en', // typically your source language in locize
supportedLngs: ['en', 'de'], // should match your available languages in locize
interpolation: {
escapeValue: false
},
react: {
useSuspense: false
}
};
// this is only used if not using suspense
i18n.options.react = yourOptions.react;
setI18n(i18n);
const backend = new LocizeBackend(backendOptions, (err, opts) => {
if (err) return console.error(err);
i18n
.use(backend)
// .use(initReactI18next) // keep this if using suspense
// yourOptions should not include backendOptions!
.init({ ...opts, ...yourOptions }, (err, t) => {
if (err) return console.error(err);
});
});
export default i18n;
Due to how serverless functions work, you cannot guarantee that a cached version of your data is available. Serverless functions are short-lived, and can shut down at any time, purging any in-memory or filesystem cache. This may be an acceptable trade-off, but sometimes it isn't acceptable.
Because of this we suggest to download the translations in your CI/CD pipeline (via cli or via api) and package them with your serverless function.
import i18next from 'i18next';
import Backend from 'i18next-fs-backend';
const backend = new Backend({
// path where resources get loaded from
loadPath: '/locales/{{lng}}/{{ns}}.json'
});
i18next
.use(backend)
.init({
// initImmediate: false, // setting initImediate to false, will load the resources synchronously
...opts,
...yourOptions
}); // yourOptions should not include backendOptions!
import i18next from 'i18next';
import en from './locales/en.json'
import de from './locales/de.json'
i18next
.init({
...opts,
...yourOptions,
resources: {
en,
de
}
});
To properly type the backend options, you can import the LocizeBackendOptions interface and use it as a generic type parameter to the i18next's init method, e.g.:
import i18n from 'i18next'
import LocizeBackend, { LocizeBackendOptions } from 'i18next-locize-backend'
i18n
.use(LocizeBackend)
.init<LocizeBackendOptions>({
backend: {
// locize backend options
},
// other i18next options
})
JavaScript
97.1%
TypeScript
1.7%
HTML
1.2%
This is an i18next backend plugin to be used for locize service. It will load resources from locize server using http fetch or xhr as fallback.
If you're not familiar with i18next and how i18next backend plugins works, please first have a look at the i18next documentation.
It will allow you to save missing keys containing both default value and context information by calling:
i18next.t(key, defaultValue, tDescription);
i18next.t(key, { defaultValue, tDescription });
To see i18next-locize-backend in a working app example, you may have a look at:
Make sure you set the debug option of i18next to true. This will maybe log more information in the developer console.
SaveMissing is not working
Did you wait 5-10 seconds before refreshing the locize UI? It may take a couple of seconds until the missing keys are sent and saved.
Per default only localhost is allowed to send missing keys (or update missing keys) (to avoid using this feature accidentally in production). If you're not using localhost during development you will have to set the allowedAddOrUpdateHosts: ['your.domain.tld'] for the backend options.
It's also recommended to set the fallbackLng equal to the source language defined in locize. i.e. if your source language in locize is de, set the fallbackLng also to de.
Loading translations not working
Make sure the translations are published, either by having enabled auto publishing for your version or by manually publishing the version. Alternatively, you can publish via CLI or directly by consuming the API.
In case you're using the private publish mode, make sure you're using the correct api key and are setting the private option to true.
import i18next from "i18next";
import Locize from "i18next-locize-backend";
i18next.use(Locize).init({
backend: {
projectId: "[PROJECTID]",
apiKey: "[APIKEY]",
version: "[VERSION]",
private: true,
referenceLng: "en",
cdnType: "standard"
},
fallbackLng: 'en', // typically your reference language in locize
supportedLngs: ['en', 'de'] // should match your available languages in locize
});
On server side: process is not exiting
In case you want to use i18next-locize-backend on server side for a short running process, you might want to set the reloadInterval option to false:
{
reloadInterval: false,
projectId: "[PROJECTID]",
version: 'latest',
referenceLng: 'en',
cdnType: 'standard'
}
On server side: downloads keep growing
On the server reloadInterval defaults to 1 hour, and the timer that drives it keeps the i18next instance alive for as long as the process runs. That is what you want for a long-lived singleton, and it is a trap for anything short-lived: if you create a new i18next instance per request or per render (a per-request instance, an SSR pass of a component that initializes i18next, a job that runs on a schedule), every one of them leaves a live timer behind and keeps refetching your namespaces forever. Nothing errors and the process still exits normally, because the timer is unrefed. The only symptom is your download count climbing between restarts.
So: create the instance once per process and reuse it, or set reloadInterval: false on instances that are not the singleton.
{
reloadInterval: false, // this instance is not a long-lived singleton
projectId: "[PROJECTID]",
version: 'latest'
}
If you are unsure whether this is happening to you, check the Downloads by Browser chart on your locize project's metrics page: server-side requests show up as i18next-locize-backend, and a per-instance leak looks like a floor that ratchets up between deploys instead of following your traffic.
Not all languages are loaded
By default the supportedLngs are defined by having a minimum of 90% of done translations. You can set a threshold for languages to be added to supportedLngs by setting translatedPercentageThreshold in backend options (eg: 1 = 100% translated, 0.9 = 90% translated).
Source can be loaded via npm, yarn, or downloaded from this repo.
# npm package
$ npm install i18next-locize-backend
# yarn
$ yarn add i18next-locize-backend
v10 requires native
fetch. Node ≥ 18, all modern browsers, Deno, and Bun shipfetchby default — no extra setup needed. v10 dropped the bundledcross-fetchfallback that v9 used. For runtimes without nativefetch, install a ponyfill yourself before loading this backend, or stay on v9.
Wiring up:
import i18next from 'i18next';
import Locize from 'i18next-locize-backend';
// or
const i18next = require('i18next');
const Locize = require('i18next-locize-backend');
i18next.use(Locize).init(i18nextOptions);
for Deno:
import i18next from 'https://deno.land/x/i18next/index.js'
import Backend from 'https://deno.land/x/i18next_locize_backend/index.js'
i18next.use(Backend).init(i18nextOptions);
window.i18nextLocizeBackendnew Locize(options, callback) constructor or the getOptions function (see below) in combination with a language detector, you should define the supportedLngs option yourself in the i18next init options. Otherwise, i18next might try to load languages that do not exist in your project, resulting in unnecessary requests.IMPORTANT make sure you do not add your apiKey in the production build to avoid misuse by strangers
{
// the id of your locize project
projectId: '[PROJECTID]',
// add an api key if you want to send missing keys
apiKey: '[APIKEY]',
// the reference language of your project
referenceLng: '[LNG]',
// version - defaults to latest
version: '[VERSION]',
// configures the cdn endpoint that should be used (depends on which cdn type you've in your locize project) - defaults to 'standard'
cdnType: 'pro',
// private - set to true if you version on locize is set to use private publish
private: false,
// hostnames that are allowed to create, update keys
// please keep those to your local system, staging, test servers (not production)
// can be array of allowed hosts or a function (hostname) => { return true; // or false if not allowed }
allowedAddOrUpdateHosts: ['localhost'],
// optional event triggered on saved to backend
onSaved: (lng, ns) => { ... },
// can be used to reload resources in a specific interval (useful in server environments)
reloadInterval: typeof window !== 'undefined' ? false : 60 * 60 * 1000,
// define the threshold for languages to be added to supportedLngs (eg: 1 = 100% translated, 0.9 = 90% translated [default]).
translatedPercentageThreshold: 0.8,
// define a custom request function
// can be used to support Angular http client
//
// 'info' contains 'url', 'method', 'body' and 'headers'
// 'url' the url that should be requested
// 'method' GET for fetching translations and POST for saving missing translations
// 'body' will be a key:value object used when saving missing translations
// 'headers' will be a key:value object containing the header information that should be sent
// 'callback' is a function that takes two parameters, 'err' and 'res'.
// 'err' should be an error
// 'res' should be an object with a 'status' property and a 'data' property containing a stringified object instance beeing the key:value translation pairs for the
// requested language and namespace, or null in case of an error.
request: function (info, callback) {},
// or async / promise
//request: async (info) {},
/**
* During development this option can be set to true, to get uncached translations.
* If not set, it will be default true, if the i18next debug option is true, else default false.
* Do NOT set this to true on production!
*/
noCache: false,
/**
* Enable a small in-memory Cache-Control-aware cache for runtimes that
* don't have a native HTTP cache (Node.js, Deno, React Native, etc).
*
* - Only affects GET requests (POST never cached).
* - Parses "Cache-Control: max-age" and stores the full response until expiry.
* - Default: true for non-browser environments, false for browsers.
*
* Note: cache is process-local and unbounded — disable for long-running apps
* that fetch many dynamic URLs.
*/
useCacheLayer: true,
}
To load translations only projectId needs to be filled. To use the saveMissing feature of i18next additional to the projectId both apiKey and referenceLng have to be set.
Options can be passed in:
preferred - by setting options.backend in i18next.init:
import i18next from "i18next";
import Locize from "i18next-locize-backend";
i18next.use(Locize).init({
backend: options,
fallbackLng: 'en', // typically your reference language in locize
supportedLngs: ['en', 'de'] // should match your available languages in locize
});
on construction:
import Locize from "i18next-locize-backend";
const locize = new Locize(options);
via calling init:
import Locize from "i18next-locize-backend";
const locize = new Locize();
locize.init(options);
Will return a list of all languages in your project including percentage of translations done per version.
import Locize from "i18next-locize-backend";
const locize = new Locize(options);
locize.getLanguages((err, data) => {
/*
data is:
{
"en": {
"name": "English",
"nativeName": "English",
"isReferenceLanguage": true,
"translated": {
"latest": 1
}
},
"de": {
"name": "German",
"nativeName": "Deutsch",
"isReferenceLanguage": false,
"translated": {
"latest": 0.9
}
}
}
*/
});
// or
const data = await locize.getLanguages();
// or
i18next.services.backendConnector.backend.getLanguages(callback);
// or
const data = await i18next.services.backendConnector.backend.getLanguages();
Will return an object containing useful informations for the i18next init options.
import Locize from "i18next-locize-backend";
const locize = new Locize(options);
locize.getOptions((err, data) => {
/*
data is:
{
fallbackLng: 'en',
referenceLng: 'en',
supportedLngs: ['en', 'de'],
load: 'languageOnly|all' // depending on your supportedLngs has locals having region like en-US
}
*/
});
// or
const data = await locize.getOptions();
// or
i18next.services.backendConnector.backend.getOptions(callback);
// or
const data = await i18next.services.backendConnector.backend.getOptions();
You can set a threshold for languages to be added to supportedLngs by setting translatedPercentageThreshold in backend options (eg: 1 = 100% translated, 0.9 = 90% translated).
You can load some information from the backend to eg. set supportedLngs for i18next just supporting languages you got in your locize project.
You will get i18next options for (same as above backend.getOptions):
import i18next from "i18next";
import Locize from "i18next-locize-backend";
const locize = new Locize(
{
projectId: "[PROJECTID]",
apiKey: "[APIKEY]",
version: "[VERSION]"
// referenceLng -> not needed as will be loaded from API
},
(err, opts, lngs) => {
i18next.use(locize).init({ ...opts, ...yourOptions }); // yourOptions should not include backendOptions!
}
);
Use setI18n to pass in the i18next instance before initializing:
import i18n from "i18next";
import { initReactI18next, setI18n } from "react-i18next";
import LocizeBackend from "i18next-locize-backend";
const backendOptions = {
projectId: "1d0aa5aa-4660-4154-b6d9-907dbef10bb3"
};
const yourOptions = {
debug: true,
fallbackLng: 'en', // typically your source language in locize
supportedLngs: ['en', 'de'], // should match your available languages in locize
interpolation: {
escapeValue: false
},
react: {
useSuspense: false
}
};
// this is only used if not using suspense
i18n.options.react = yourOptions.react;
setI18n(i18n);
const backend = new LocizeBackend(backendOptions, (err, opts) => {
if (err) return console.error(err);
i18n
.use(backend)
// .use(initReactI18next) // keep this if using suspense
// yourOptions should not include backendOptions!
.init({ ...opts, ...yourOptions }, (err, t) => {
if (err) return console.error(err);
});
});
export default i18n;
Due to how serverless functions work, you cannot guarantee that a cached version of your data is available. Serverless functions are short-lived, and can shut down at any time, purging any in-memory or filesystem cache. This may be an acceptable trade-off, but sometimes it isn't acceptable.
Because of this we suggest to download the translations in your CI/CD pipeline (via cli or via api) and package them with your serverless function.
import i18next from 'i18next';
import Backend from 'i18next-fs-backend';
const backend = new Backend({
// path where resources get loaded from
loadPath: '/locales/{{lng}}/{{ns}}.json'
});
i18next
.use(backend)
.init({
// initImmediate: false, // setting initImediate to false, will load the resources synchronously
...opts,
...yourOptions
}); // yourOptions should not include backendOptions!
import i18next from 'i18next';
import en from './locales/en.json'
import de from './locales/de.json'
i18next
.init({
...opts,
...yourOptions,
resources: {
en,
de
}
});
To properly type the backend options, you can import the LocizeBackendOptions interface and use it as a generic type parameter to the i18next's init method, e.g.:
import i18n from 'i18next'
import LocizeBackend, { LocizeBackendOptions } from 'i18next-locize-backend'
i18n
.use(LocizeBackend)
.init<LocizeBackendOptions>({
backend: {
// locize backend options
},
// other i18next options
})
JavaScript
97.1%
TypeScript
1.7%
HTML
1.2%