General sensenet independent client side utilities

Usage no npm install needed!

<script type="module">
  import sensenetClientUtils from 'https://cdn.skypack.dev/@sensenet/client-utils';



General sensenet independent client side utilities

NPM version NPM downloads License: GPL v2


# Yarn
yarn add @sensenet/client-utils

npm install @sensenet/client-utils


You can implement disposable resources and use them with a using() or usingAsync() syntax. Example:

class Resource implements Disposable {
  dispose() {
    // cleanup logics

using(new Resource(), resource => {
  // do something with the resource

usingAsync(new Resource(), async resource => {
  // do something with the resource, allows awaiting promises

ObservableValue and ValueObservers

You can track value changes using with this simple Observable implementation.


const observableValue = new ObservableValue<number>(0)
const observer = observableValue.subscribe(newValue => {
  console.log('Value changed:', newValue)

// To update the value
// if you want to dispose a single observer
// if you want to dispose the whole observableValue with all of its observers:


The class contains small helper methods for path transformation and manipulation.


Retrier is a utility that can keep trying an operation until it succeeds, times out or reach a specified retry limit.

const funcToRetry: () => Promise<boolean> = async () => {
  const hasSucceeded = false
  // ...
  // custom logic
  // ...
  return hasSucceeded
const retrierSuccess = await Retrier.create(funcToRetry)
    Retries: 3,
    RetryIntervalMs: 1,
    timeoutMs: 1000,


Trace is an utility that can be used to track method calls, method returns and errors

const methodTracer: Disposable = Trace.method({
  object: myObjectInstance, // You can define an object constructor for static methods as well
  method: myObjectInstance.method, // The method to be tracked
  methodName: 'method', // Unique identifier for the method
  isAsync: true, // if you set to async, method finished will be *await*-ed
  onCalled: traceData => {
    console.log('Method called:', traceData)
  onFinished: traceData => {
    console.log('Method call finished:', traceData)
  onError: traceData => {
    console.log('Method throwed an error:', traceData)

// if you want to stop receiving events

Logging package

Initializing with @sensenet-client-utils/inject

You can start using the Logging service with an injector in the following way

import { ConsoleLogger, Injector } from '@sensenet/client-utils'

const myInjector = new Injector().useLogging(ConsoleLogger, Logger1, Logger2 /** ...your Logger implementations */)

You can retrieve the Logger instance with

const myLogger = myInjector.logger

Logging events

You can log a simple event with

  level: LogLevel.Verbose,
  message: 'My log message',
  scope: 'logger/test',
  data: {
    foo: 1,
    bar: 42,


  message: 'My log message',
  scope: '@sensenet/client-utils/test',
  data: {
    foo: 1,
    bar: 42,

The two snippets do the same - they will add a log entry to each registered logger.

Scoped loggers

At the most of the cases, you use a logger in a service with a specific scope. You can create and use a scoped logger in the following way

const scopedLogger = myLogger.withScope('logger/test')
scopedLogger.verbose({ message: 'FooBarBaz' })
Implementing your own logger

You can implement your own logging logic in the similar way as this custom log collector

import { AbstractLogger, Injectable, LeveledLogEntry } from '@sensenet/client-utils'

@Injectable({ lifetime: 'singleton' })
export class MyCustomLogCollector extends AbstractLogger {
  private readonly entries: Array<LeveledLogEntry<any>> = []

  public getEntries() {
    return [...this.entries]

  public async addEntry<T>(entry: LeveledLogEntry<T>): Promise<void> {

  constructor() {


Injectors act as containers, they are responsible for creating / retrieving service instances based on the provided Injectable metadata. You can create an injector with simply instantiating the class

const myInjector = new Injector()

You can organize your injector(s) in trees by creating child injectors. You can use the children and services with scoped lifetime for contextual services.

const childInjector = myInjector.createChild({ owner: 'myCustomContext' })


Creating an Injectable service from a class

You can create an injectable service from a plain class when decorating with the @Injectable() decorator.

  /** Injectable options */
export class MySercive {
  /** ...service implementation... */

  constructor(s1: OtherInjectableService, s2: AnotherInjectableService) {}

The constructor parameters (s1: OtherInjectableService and s2: AnotherInjectableService) should be also decorated and will be resolved recursively.


You can define a specific Lifetime for Injectable services on the decorator

  lifetime: 'transient',
export class MySercive {
  /** ...service implementation... */

The lifetime can be

  • transient - A new instance will be created each time when you get an instance
  • scoped - A new instance will be created if it doesn't exist on the current scope. Can be useful for injectable services that can be used for contextual data.
  • singleton - A new instance will be created only if it doesn't exists on the root injector. It will act as a singleton in other cases.

Injectables can only depend on services with longer lifetime, e.g. a transient can depend on a singleton, but inversing it will throw an error

Retrieving your service from the injector

You can retrieve a service by calling

const service = myInjector.getInstance(MySercive)
Explicit instance setup

There are cases that you have to set a service instance explicitly. You can do that in the following way

class MyService {
  constructor(public readonly foo: string)

myInjector.setExplicitInstance(new MyService('bar'))