@ale-rainbow/i18n
Advanced tools
lightweight translation module with dynamic json storage, forked from original i18n-node module
Lightweight simple translation module with dynamic json storage. Supports plain vanilla node.js apps and should work with any framework (like express, restify and probably more) that exposes an app.use() method passing in res and req objects.
Uses common __('...') syntax in app and templates.
Stores language files in json files compatible to webtranslateit json format.
Adds new strings on-the-fly when first used in your app.
No extra parsing needed.
npm install i18n --save
npm test
// load modules
var express = require('express'),
i18n = require("i18n");
Minimal example, just setup two locales and a project specific directory
i18n.configure({
locales:['en', 'de'],
directory: __dirname + '/locales'
});
now you are ready to use a global i18n.__('Hello').
In your cli, when not registered to a specific object:
var greeting = i18n.__('Hello');
Global assumes you share a common state of localization in any time and any part of your app. This is usually fine in cli-style scripts. When serving responses to http requests you'll need to make sure that scope is NOT shared globally but attached to your request object.
In an express app, you might use i18n.init to gather language settings of your visitors and also bind your helpers to response object honoring request objects locale, ie:
// Configuration
app.configure(function() {
[...]
// default: using 'accept-language' header to guess language settings
app.use(i18n.init);
[...]
});
in your apps methods:
app.get('/de', function(req, res){
var greeting = res.__('Hello');
});
in your templates (depending on your template engine)
<%= __('Hello') %>
${__('Hello')}
See tested examples inside /examples for some inspiration in node 4.x / 5.x or browse these gists:
PLEASE NOTE: Those gist examples worked until node 0.12.x only
For serving the same static files with different language url, you could:
app.use(express.static(__dirname + '/www'));
app.use('/en', express.static(__dirname + '/www'));
app.use('/de', express.static(__dirname + '/www'));
The api is subject of incremental development. That means, it should not change nor remove any aspect of the current api but new features and options will get added that don't break compatibility backwards within a major version.
You should configure your application once to bootstrap all aspects of i18n. You should not configure i18n in each loop when used in an http based scenario. During configuration, i18n reads all known locales into memory and prepares to keep that superfast object in sync with your files in filesystem as configured
i18n.configure({
locales:['en', 'de'],
directory: __dirname + '/locales'
});
Since 0.7.0 you may even omit the locales setting and just configure a directory. i18n will read all files within that directory and detect all given locales by their filenames.
i18n.configure({
directory: __dirname + '/locales'
});
i18n.configure({
// setup some locales - other locales default to en silently
locales:['en', 'de'],
// fall back from Dutch to German
fallbacks:{'nl': 'de'},
// you may alter a site wide default locale
defaultLocale: 'de',
// sets a custom cookie name to parse locale settings from - defaults to NULL
cookie: 'yourcookiename',
// query parameter to switch locale (ie. /home?lang=ch) - defaults to NULL
queryParameter: 'lang',
// where to store json files - defaults to './locales' relative to modules directory
directory: './mylocales',
// controll mode on directory creation - defaults to NULL which defaults to umask of process user. Setting has no effect on win.
directoryPermissions: '755',
// watch for changes in json files to reload locale on updates - defaults to false
autoReload: true,
// whether to write new locale information to disk - defaults to true
updateFiles: false,
// sync locale information across all files - defaults to false
syncFiles: false,
// what to use as the indentation unit - defaults to "\t"
indent: "\t",
// setting extension of json files - defaults to '.json' (you might want to set this to '.js' according to webtranslateit)
extension: '.js',
// setting prefix of json files name - default to none '' (in case you use different locale files naming scheme (webapp-en.json), rather then just en.json)
prefix: 'webapp-',
// enable object notation
objectNotation: false,
// setting of log level DEBUG - default to require('debug')('i18n:debug')
logDebugFn: function (msg) {
console.log('debug', msg);
},
// setting of log level WARN - default to require('debug')('i18n:warn')
logWarnFn: function (msg) {
console.log('warn', msg);
},
// setting of log level ERROR - default to require('debug')('i18n:error')
logErrorFn: function (msg) {
console.log('error', msg);
},
// object or [obj1, obj2] to bind the i18n api and current locale to - defaults to null
register: global,
// hash to specify different aliases for i18n's internal methods to apply on the request/response objects (method -> alias).
// note that this will *not* overwrite existing properties with the same name
api: {
'__': 't', //now req.__ becomes req.t
'__n': 'tn' //and req.__n can be called as req.tn
},
// Downcase locale when passed on queryParam; e.g. lang=en-US becomes
// en-us. When set to false, the queryParam value will be used as passed;
// e.g. lang=en-US remains en-US.
preserveLegacyCase: true
});
The locale itself is gathered directly from the browser by header, cookie or query parameter depending on your setup.
In case of cookie you will also need to enable cookies for your application. For express this done by adding app.use(express.cookieParser())). Now use the same cookie name when setting it in the user preferred language, like here:
res.cookie('yourcookiename', 'de', { maxAge: 900000, httpOnly: true });
After this and until the cookie expires, i18n.init() will get the value of the cookie to set that language instead of default for every page.
register optionEsp. when used in a cli like scriptyou won't use any i18n.init() to guess language settings from your user. Thus i18n won't bind itself to any res or req object and will work like a static module.
var anyObject = {};
i18n.configure({
locales: ['en', 'de'],
register: anyObject
});
anyObject.setLocale('de');
anyObject.__('Hallo'); // --> Hallo`
Cli usage is a special use case, as we won't need to maintain any transaction / concurrency aware setting of locale, so you could even choose to bind i18n to global scope of node:
i18n.configure({
locales: ['en', 'de'],
register: global
});
i18n.setLocale('de');
__('Hello'); // --> Hallo`
When used as middleware in frameworks like express to setup the current environment for each loop. In contrast to configure the i18n.init() should be called within each request-response-cycle.
var app = express();
app.use(cookieParser());
app.use(i18n.init);
When i18n is used like this, the i18n.init() tries to
Express would call i18n.init(req, res, next), which is "classic" and adopted by many frameworks. Thus i18n will attach it's api to that schema:
{
req: {
locals: {},
res: {
locals: {},
}
}
}
and add it's extra attributes and methods, like so:
{
req: {
locals: {
locale: "de",
__: [function],
__n: [function],
[...]
},
res: {
locals: {
locale: "de",
__: [function],
__n: [function],
[...]
},
locale: "de",
__: [function],
__n: [function],
[...]
},
locale: "de",
__: [function],
__n: [function],
[...]
}
}
Now each local object (ie. res.locals) is setup with it's own "private" locale and methods to get the appropriate translation from the global catalog.
Translates a single phrase and adds it to locales if unknown. Returns translated parsed and substituted string.
// template and global (this.locale == 'de')
__('Hello'); // Hallo
__('Hello %s', 'Marcus'); // Hallo Marcus
__('Hello {{name}}', { name: 'Marcus' }); // Hallo Marcus
// scoped via req object (req.locale == 'de')
req.__('Hello'); // Hallo
req.__('Hello %s', 'Marcus'); // Hallo Marcus
req.__('Hello {{name}}', { name: 'Marcus' }); // Hallo Marcus
// scoped via res object (res.locale == 'de')
res.__('Hello'); // Hallo
res.__('Hello %s', 'Marcus'); // Hallo Marcus
res.__('Hello {{name}}', { name: 'Marcus' }); // Hallo Marcus
// passing specific locale
__({phrase: 'Hello', locale: 'fr'}); // Salut
__({phrase: 'Hello %s', locale: 'fr'}, 'Marcus'); // Salut Marcus
__({phrase: 'Hello {{name}}', locale: 'fr'}, { name: 'Marcus' }); // Salut Marcus
Plurals translation of a single phrase. Singular and plural forms will get added to locales if unknown. Returns translated parsed and substituted string based on last count parameter.
// short syntax is best suited for reading
// --> writes '%s cat' to both `one` and `other` plurals
__n('%s cat', 1) // --> 1 Katze
__n('%s cat', 3) // --> 3 Katzen
// long syntax works fine in combination with `updateFiles`
// --> writes '%s cat' to `one` and '%s cats' to `other` plurals
// "one" (singular) & "other" (plural) just covers the basic Germanic Rule#1 correctly.
__n("%s cat", "%s cats", 1); // 1 Katze
__n("%s cat", "%s cats", 3); // 3 Katzen
// scoped via req object (req.locale == 'de')
req.__n("%s cat", 1); // 1 Katze
req.__n("%s cat", 3); // 3 Katzen
// scoped via res object (res.locale == 'de')
res.__n("%s cat", 1); // 1 Katze
res.__n("%s cat", 3); // 3 Katzen
// passing specific locale
__n({singular: "%s cat", plural: "%s cats", locale: "fr"}, 1); // 1 chat
__n({singular: "%s cat", plural: "%s cats", locale: "fr"}, 3); // 3 chats
// the all in one object signature
__n({singular: "%s cat", plural: "%s cats", locale: "fr", count: 1}); // 1 chat
__n({singular: "%s cat", plural: "%s cats", locale: "fr", count: 3}); // 3 chats
When used in short form like __n(phrase, count) the following will get added to your json files:
__n('%s dog', 1)
{
"%s dog": {
"one": "%s dog",
"other": "%s dog"
}
}
When used in long form like __n(singular, plural, count) you benefit form passing defaults to both forms:
__n('%s kitty', '%s kittens', 0)
{
"%s kitty": {
"one": "%s kitty",
"other": "%s kittens"
}
}
You might now add extra forms to certain json files to support the complete set of plural forms, like for example in russian:
{
"%s cat": {
"one": "%d кошка",
"few": "%d кошки",
"many": "%d кошек",
"other": "%d кошка",
}
}
and let __n() select the correct form for you:
__n('%s cat', 0); // --> 0 кошек
__n('%s cat', 1); // --> 1 кошка
__n('%s cat', 2); // --> 2 кошки
__n('%s cat', 5); // --> 5 кошек
__n('%s cat', 6); // --> 6 кошек
__n('%s cat', 21); // --> 21 кошка
Note i18n.__n() will add a blueprint ("one, other" or "one, few, other" for eaxmple) for each locale to your json on updateFiles in a future version.
Supports the advanced MessageFormat as provided by excellent messageformat module. You should definetly head over to messageformat.github.io for a guide to MessageFormat. i18n takes care of new MessageFormat('en').compile(msg); with the current msg loaded from it's json files and cache that complied fn in memory. So in short you might use it similar to __() plus extra object to accomblish MessageFormat's formating. Ok, some examples:
// assume res is set to german
res.setLocale('de');
// start simple
res.__mf('Hello'); // --> Hallo
// can replace too
res.__mf('Hello {name}', { name: 'Marcus' }) // --> Hallo Marcus
// and combines with sprintf
res.__mf('Hello {name}, how was your %s?', 'test', { name: 'Marcus' }) // --> Hallo Marcus, wie war dein test?
// now check out a plural rule
res.__mf('{N, plural, one{# cat} few{# cats} many{# cats} others{# cats}}', {N: 1})
// results for "1" in (all use "one")
// en --> 1 cat
// de --> 1 Katze
// fr --> 1 chat
// ru --> 1 кошка ru uses "__one__" when ending on "1"
// results for "0" in (most use "others")
// en --> 0 cats
// de --> 0 Katzen
// fr --> 0 chat fr uses "__one__" for zero
// ru --> 0 кошек ru uses "__many__"
// results for "2" in (most use "others")
// en --> 2 cat
// de --> 2 Katze
// fr --> 2 chat
// ru --> 2 кошки ru uses "__few__" when ending on "1"
// results for "5" in (most use "others")
// en --> 5 cat
// de --> 5 Katze
// fr --> 5 chat
// ru --> 5 кошек ru uses "__many__"
// results for "21" in (most use "others")
// en --> 21 cat
// de --> 21 Katze
// fr --> 21 chat
// ru --> 21 кошка ru uses "__one__" when ending on "1"
Take a look at Mozilla to quickly get an idea of what pluralization has to deal with. With __mf() you get a very powerfull tool, but you need to handle it correctly.
But MessageFormat can handle more! You get ability to process:
Combinations of those give superpower, but should get tested well (contribute your use case, please!) on integration.
Returns a list of translations for a given phrase in each language.
i18n.__l('Hello'); // --> [ 'Hallo', 'Hello' ]
This will be usefull when setting up localized routes for example (kudos to @xpepermint, #150):
// this will match routes
// EN --> /:locale/products/:id?
// ES --> /:locale/productos/:id?
app.get( __l('/:locale/products/:id?'), function (req, res) {
// guess what you might use req.params.locale for?
});
i18n.__ln() to get plurals will come up in another release...
Returns a hashed list of translations for a given phrase in each language.
i18n.__h('Hello'); // --> [ { de: 'Hallo' }, { en: 'Hello' } ]
i18n.__hn() to get plurals will come up in another release...
Setting the current locale (ie.: en) globally or in current scope.
setLocale('de');
setLocale(req, 'de');
req.setLocale('de');
Use setLocale to change any initial locale that was set in i18n.init(). You get more control on how when and which objects get setup with a given locale. Locale values are inherited within the given schema like in i18n.init() Let's see some examples:
i18n.setLocale(req, 'ar'); // --> req: مرحبا res: مرحبا res.locals: مرحبا
i18n.setLocale(res, 'ar'); // --> req: Hallo res: مرحبا res.locals: مرحبا
i18n.setLocale(res.locals, 'ar'); // --> req: Hallo res: Hallo res.locals: مرحبا
You'll get even more controll when passing an array of objects:
i18n.setLocale([req, res.locals], req.params.lang); // --> req: مرحبا res: Hallo res.locals: مرحبا
or disable inheritance by passing true as third parameter:
i18n.setLocale(res, 'ar', true); // --> req: Hallo res: مرحبا res.locals: Hallo
Getting the current locale (ie.: en) from current scope or globally.
getLocale(); // --> de
getLocale(req); // --> de
req.getLocale(); // --> de
Returns a list with all configured locales.
i18n.getLocales(); // --> ['en', 'de', 'en-GB']
Returns a whole catalog optionally based on current scope and locale.
getCatalog(); // returns catalog for all locales
getCatalog('de'); // returns just for 'de'
getCatalog(req); // returns catalog for all locales
getCatalog(req, 'de'); // returns just for 'de'
req.getCatalog(); // returns catalog for all locales
req.getCatalog('de'); // returns just for 'de'
In general i18n has to be attached to the response object to let it's public api get accessible in your templates and methods. As of 0.4.0 i18n tries to do so internally via i18n.init, as if you were doing it in app.configure on your own:
app.use(function(req, res, next) {
// express helper for natively supported engines
res.locals.__ = res.__ = function() {
return i18n.__.apply(req, arguments);
};
[...]
next();
});
Different engines need different implementations, so yours might miss or not work with the current default helpers. This one showing an example for mustache in express:
// register helper as a locals function wrapped as mustache expects
app.use(function (req, res, next) {
// mustache helper
res.locals.__ = function () {
return function (text, render) {
return i18n.__.apply(req, arguments);
};
};
[...]
next();
});
You could still setup your own implementation. Please refer to Examples below, post an issue or contribute your setup.
As inspired by gettext there is currently support for sprintf-style expressions. You can also use mustache syntax for named parameters.
var greeting = __('Hello %s, how are you today?', 'Marcus');
this puts Hello Marcus, how are you today?. You might add endless arguments and even nest it.
var greeting = __('Hello %s, how are you today? How was your %s.', 'Marcus', __('weekend'));
which puts Hello Marcus, how are you today? How was your weekend.
You might need to have repeated references to the same argument, which can be done with sprintf.
var example = __('%1$s, %1$s, %1$s', 'repeat');
which puts
repeat, repeat, repeat
In some cases the argument order will need to be switched for different locales. The arguments can be strings, floats, numbers, etc.
var example = __('%2$d then %1$s then %3$.2f', 'First', 2, 333.333);
which puts
2 then First then 333.33
You may also use mustache syntax for your message strings. To pass named parameters to your message, just provide an object as the last parameter. You can still pass unnamed parameters by adding additional arguments.
var greeting = __('Hello {{name}}, how are you today?', { name: 'Marcus' });
this puts Hello Marcus, how are you today?. You might also combine it with sprintf arguments...
var greeting = __('Hello {{name}}, how was your %s.', __('weekend'), { name: 'Marcus' });
and even nest it...
var greeting = __( __('Hello {{name}}, how was your %s?', { name: 'Marcus' }), __('weekend') );
which both put Hello Marcus, how was your weekend.
two different plural forms are supported as response to count:
var singular = __n('%s cat', '%s cats', 1);
var plural = __n('%s cat', '%s cats', 3);
this puts 1 cat or 3 cats and again these could get nested:
var singular = __n('There is one monkey in the %%s', 'There are %d monkeys in the %%s', 1, 'tree');
var plural = __n('There is one monkey in the %%s', 'There are %d monkeys in the %%s', 3, 'tree');
putting There is one monkey in the tree or There are 3 monkeys in the tree. Passing all 3 parameters would write a one and other to your json. For reading you might just use 2 parameters, too:
__n('%s cat', 1) // --> 1 Katze
__n('%s cat', 3) // --> 3 Katzen
use mathematical intervals to declare any own plural rules based on ISO 31-11 notation. Let's assume the following json snippet:
"dogs": {
"one": "one dog",
"other": "[0] no dog|[2,5] some dogs|[6,11] many dogs|[12,36] dozens of dogs|a horde of %s dogs|[100,] too many dogs"
}
this will result in
__n('dogs', 0) // --> no dog
__n('dogs', 1) // --> one dog
__n('dogs', 2) // --> some dogs
__n('dogs', 10) // --> many dogs
__n('dogs', 25) // --> dozens of dogs
__n('dogs', 42) // --> a horde of 42 dogs
__n('dogs', 199) // --> too many dogs
The rules are parsed in sequenced order, so the first match will skip any extra rules. Example:
{
"dogs":"[0]no dog|[1]one dog|[,10[ less than ten dogs|[,20[ less than 20 dogs|too many dogs"
}
results in
// interval matched by number
__n('dogs', 0) // --> no dog
__n('dogs', 1) // --> one dog
__n('dogs', 2) // --> less than ten dogs
__n('dogs', 9) // --> less than ten dogs
__n('dogs', 10) // --> less than 20 dogs
__n('dogs', 19) // --> less than 20 dogs
__n('dogs', 20) // --> too many dogs
__n('dogs', 199) // --> too many dogs
// no interval returned, but found a catchall
__('dogs') // --> too many dogs
See en.json example inside /locales for some inspiration on use cases. Each phrase might get decorated further with mustache and sprintf expressions:
{
"example":"[0] %s is zero rule for {{me}}|[2,5] %s is between two and five for {{me}}|and a catchall rule for {{me}} to get my number %s"
}
will put (as taken from tests):
// will always use a found catchall
__('example', {me: 'marcus'}) // --> and a catchall rule for marcus to get my number %s
__('example', ['one'], {me: 'marcus'}) // --> and a catchall rule for marcus to get my number one
// will search a matching interval or fallback to catchall
__n('example', 1, {me: 'marcus'}) // --> and a catchall rule for marcus to get my number 1
__n('example', 2, {me: 'marcus'}) // --> 2 is between two and five for marcus
__n('example', 5, {me: 'marcus'}) // --> 5 is between two and five for marcus
__n('example', 3, {me: 'marcus'}) // --> 3 is between two and five for marcus
__n('example', 6, {me: 'marcus'}) // --> and a catchall rule for marcus to get my number 6
Notice: the "example" object in your json doesn't use any "one", "other" subnodes although you could use and even combine them. Currently "one" referres to the value of exactly 1 while "other" referres to every other value (think of 0, -10, null, false)
you might even use dynamic variables as they get interpreted on the fly. Better make sure no user input finds it's way to that point as they all get added to the en.js file if not yet existing.
var greetings = ['Hi', 'Hello', 'Howdy'];
for (var i=0; i < greetings.length; i++) {
console.log( __(greetings[i]) );
};
which puts
Hi
Hello
Howdy
In addition to the traditional, linear translation lists, i18n also supports hierarchical translation catalogs.
To enable this feature, be sure to set objectNotation to true in your configure() call. Note: If you can't or don't want to use . as a delimiter, set objectNotation to any other delimiter you like.
Instead of calling __("Hello") you might call __("greeting.formal") to retrieve a formal greeting from a translation document like this one:
"greeting": {
"formal": "Hello",
"informal": "Hi",
"placeholder": {
"formal": "Hello %s",
"informal": "Hi %s"
}
}
In the document, the translation terms, which include placeholders, are nested inside the "greeting" translation. They can be accessed and used in the same way, like so __('greeting.placeholder.informal', 'Marcus').
Object notation also supports pluralization. When making use of it, the "one" and "other" entries are used implicitly for an object in the translation document. For example, consider the following document:
"pets":{
"cat": {
"one": "Katze",
"other": "Katzen"
}
}
When accessing these, you would use __n("pets.cat", "pets.cat", 3) to tell i18n to use both the singular and plural form of the "cat" entry. Naturally, you could also access these members explicitly with __("pets.cat.one") and __("pets.cat.other").
When starting a project from scratch, your translation documents will probably be empty. i18n takes care of filling your translation documents for you. Whenever you use an unknown object, it is added to the translation documents.
By default, when using object notation, the provided string literal will be inserted and returned as the default string. As an example, this is what the "greeting" object shown earlier would look like by default:
"greeting": {
"formal": "greeting.formal",
"informal": "greeting.informal"
}
In case you would prefer to have a default string automatically inserted and returned, you can provide that default string by appending it to your object literal, delimited by a :. For example:
__("greeting.formal:Hello")
__("greeting.placeholder.informal:Hi %s")
Will get modular support for different storage engines, currently just json files are stored in filesystem.
the above will automatically generate a en.json by default inside ./locales/ which looks like
{
"Hello": "Hello",
"Hello %s, how are you today?": "Hello %s, how are you today?",
"weekend": "weekend",
"Hello %s, how are you today? How was your %s.": "Hello %s, how are you today? How was your %s.",
"Hi": "Hi",
"Howdy": "Howdy",
"%s cat": {
"one": "%s cat",
"other": "%s cats"
},
"There is one monkey in the %%s": {
"one": "There is one monkey in the %%s",
"other": "There are %d monkeys in the %%s"
},
"tree": "tree",
"%s dog": {
"one": "one dog",
"other": "[0] no dog|[2,5] some dogs|[6,11] many dogs|[12,36] dozens of dogs|a horde of %s dogs"
}
}
that file can be edited or just uploaded to webtranslateit for any kind of collaborative translation workflow:
{
"Hello": "Hallo",
"Hello %s, how are you today?": "Hallo %s, wie geht es dir heute?",
"weekend": "Wochenende",
"Hello %s, how are you today? How was your %s.": "Hallo %s, wie geht es dir heute? Wie war dein %s.",
"Hi": "Hi",
"Howdy": "Hallöchen",
"%s cat": {
"one": "%s Katze",
"other": "%s Katzen"
},
"There is one monkey in the %%s": {
"one": "Im %%s sitzt ein Affe",
"other": "Im %%s sitzen %d Affen"
},
"tree": "Baum",
"%s dog": {
"one": "Ein Hund",
"other": "[0] Kein Hund|[2,5] Ein paar Hunde|[6,11] Viele Hunde|[12,36] Dutzende Hunde|Ein Rudel von %s Hunden"
}
}
Logging any kind of output is moved to debug module. To let i18n output anything run your app with DEBUG env set like so:
$ DEBUG=i18n:* node app.js
i18n exposes three log-levels:
if you only want to get errors and warnings reported start your node server like so:
$ DEBUG=i18n:warn,i18n:error node app.js
Combine those settings with you existing application if any of you other modules or libs also uses debug
You can configure i18n to use a custom logger. For example attach some simple console-logging:
i18n.configure({
...
// setting of log level DEBUG - default to require('debug')('i18n:debug')
logDebugFn: function (msg) {
console.log('debug', msg);
},
// setting of log level WARN - default to require('debug')('i18n:warn')
logWarnFn: function (msg) {
console.log('warn', msg);
},
// setting of log level ERROR - default to require('debug')('i18n:error')
logErrorFn: function (msg) {
console.log('error', msg);
}
});
i18n.setLocale() recursion bug on nested res-/req-objects sails#3631i18n.__n() supports all pluralsi18n.__mf(), api alias option, syncFiles optioni18n.setLocale() and i18n.init() refactored to comply with most common use cases, much better test coverage and docsautoReload, directoryPermissions, register, queryParameter, read locales from filenames with empty locales option (#134)i18n.setLocale()i18n.getLocales(), custom logger, fallback[s];i18n.setLocale() for req and resi18n.getCatalogthis.locale to __ and __nFAQs
lightweight translation module with dynamic json storage, forked from original i18n-node module
We found that @ale-rainbow/i18n demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 6 open source maintainers collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
Research
A malicious npm package disguised as a WhatsApp client is exploiting authentication flows with a remote kill switch to exfiltrate data and destroy files.
Security News
PyPI now supports digital attestations, enhancing security and trust by allowing package maintainers to verify the authenticity of Python packages.
Security News
GitHub removed 27 malicious pull requests attempting to inject harmful code across multiple open source repositories, in another round of low-effort attacks.