@handy-common-utils/promise-utils
Advanced tools
These Promise related utilities have 100% test coverage. The package is tiny because there is no dependency on any other package.
Functions provided are repeat, withRetry, inParallel, delayedResolve, delayedReject, timoutResolve, timeoutReject, promiseState, synchronized, etc.
First add it as a dependency:
npm install @handy-common-utils/promise-utils
Then you can use it in the code:
import { PromiseUtils } from '@handy-common-utils/promise-utils';
// delayedResolve(...), delayedReject(...), promiseState(...)
const p1 = PromiseUtils.delayedResolve(50, 1);
const p2 = PromiseUtils.delayedReject(50, 2);
await expect(PromiseUtils.promiseState(p1)).eventually.eq(PromiseState.Pending);
await expect(PromiseUtils.promiseState(p2)).eventually.eq(PromiseState.Pending);
await PromiseUtils.delayedResolve(80);
await expect(PromiseUtils.promiseState(p1)).eventually.eq(PromiseState.Fulfilled);
await expect(PromiseUtils.promiseState(p2)).eventually.eq(PromiseState.Rejected);
// timeoutReject(...)
const p = PromiseUtils.timeoutReject(PromiseUtils.delayedReject(80, '1'), 10, '2');
await expect(p).to.be.rejectedWith('2');
// repeat(...)
async repeatFetchingItemsByPosition<T>(
fetchItemsByPosition: (parameter: { position?: string }) => Promise<{ position?: string; items?: Array<T> }>,
) {
return PromiseUtils.repeat(
fetchItemsByPosition,
response => response.position ? { position: response.position } : null,
(collection, response) => response.items ? collection.concat(response.items) : collection,
[] as Array<T>,
);
}
You can either import and use the PromiseUtils class as shown above, or you can import its re-exported functions directly like below:
import { withRetry, inParallel } from '@handy-common-utils/promise-utils';
// withRetry(...)
const result = await PromiseUtils.withRetry(() => doSomething(), [100, 200, 300, 500, 800, 1000]);
const result2 = await PromiseUtils.withRetry(() => doSomething(), PromiseUtils.FIBONACCI_SEQUENCE, err => err.statusCode === 429);
// inParallel(...)
const topicArns = topics.map(topic => topic.TopicArn!);
await inParallel(5, topicArns, async topicArn => {
const topicAttributes = (await sns.getTopicAttributes({ TopicArn: topicArn }).promise()).Attributes!;
const topicDetails = { ...topicAttributes, subscriptions: [] } as any;
if (this.shouldInclude(topicArn)) {
inventory.snsTopicsByArn.set(topicArn, topicDetails);
}
});
@handy-common-utils/promise-utils
• Const FIBONACCI_SEQUENCE: number[]
Array of Fibonacci numbers starting from 1 up to 317811.
@handy-common-utils/promise-utils / PromiseUtils
• new PromiseUtils()
▸ Static delayedReject<T, R>(ms, reason): Promise<T>
Create a Promise that rejects after number of milliseconds specified.
| Name | Type |
|---|---|
T | never |
R | any |
| Name | Type | Description |
|---|---|---|
ms | number | number of milliseconds after which the created Promise would reject |
reason | R | PromiseLike<R> | () => R | PromiseLike<R> | the reason of the rejection for the Promise, or a function that supplies the reason. If the reason ends up to be a rejected Promise, then the outcome (could be fulfilled or rejected) of it will be the reject reason of the Promise returned. |
Promise<T>
the new Promise created
▸ Static delayedResolve<T>(ms, result?): Promise<T>
Create a Promise that resolves after number of milliseconds specified
| Name |
|---|
T |
| Name | Type | Description |
|---|---|---|
ms | number | number of milliseconds after which the created Promise would resolve |
result? | T | PromiseLike<T> | () => T | PromiseLike<T> | the result to be resolved for the Promise, or a function that supplies the reuslt. |
Promise<T>
the new Promise created
▸ Static inParallel<Data, Result, TError>(parallelism, jobs, operation): Promise<(Result | TError)[]>
Run multiple jobs/operations in parallel.
example
const topicArns = topics.map(topic => topic.TopicArn!);
await PromiseUtils.inParallel(5, topicArns, async topicArn => {
const topicAttributes = (await sns.getTopicAttributes({ TopicArn: topicArn }).promise()).Attributes!;
const topicDetails = { ...topicAttributes, subscriptions: [] } as any;
if (this.shouldInclude(topicArn)) {
inventory.snsTopicsByArn.set(topicArn, topicDetails);
}
});
| Name | Type | Description |
|---|---|---|
Data | Data | Type of the job data, usually it would be an Array |
Result | Result | Type of the return value of the operation function |
TError | Result | - |
| Name | Type | Description |
|---|---|---|
parallelism | number | how many jobs/operations can be running at the same time |
jobs | Iterable<Data> | job data which will be the input to operation function. This function is safe when there are infinite unknown number of elements in the job data. |
operation | (job: Data, index: number) => Promise<Result> | the function that turns job data into result asynchronously |
Promise<(Result | TError)[]>
Promise of void if the operation function does not return a value, or promise of an array containing results returned from the operation function. In the array containing results, each element is either the fulfilled result, or the rejected error/reason.
▸ Static promiseState(p): Promise<PromiseState>
Get the state of the Promise. Please note that the returned value is a Promise, although it resolves immediately.
| Name | Type | Description |
|---|---|---|
p | Promise<any> | the Promise for which we would like to know its state |
Promise<PromiseState>
A Promise that resolves immediately cotaining the state of the input Promise
▸ Static repeat<Result, Param, Collection>(operation, nextParameter, collect, initialCollection, initialParameter?): Promise<Collection>
Do an operation repeatedly and collect all the results. This function is useful for client side pagination.
example
const domainNameObjects = await PromiseUtils.repeat(
pagingParam => apig.getDomainNames({limit: 500, ...pagingParam}).promise(),
esponse => response.position? {position: response.position} : null,
(collection, response) => collection.concat(response.items!),
[] as APIGateway.DomainName[],
);
| Name | Description |
|---|---|
Result | type of the operation result |
Param | type of the input to the operation, normally the input is a paging parameter |
Collection | type of the returned value of this function |
| Name | Type | Description |
|---|---|---|
operation | (parameter: Partial<Param>) => Promise<Result> | a function that takes paging parameter as input and outputs a result, normally the operation supports paging |
nextParameter | (response: Result) => null | Partial<Param> | Promise<Partial<Param>> | The function for calculating next parameter from the operation result. Normally the parameter controls paging, This function should return null when next invocation of the operation function is not desired. If next invocation is desired, the return value of this function can be a Promise or not a Promise. |
collect | (collection: Collection, result: Result) => Collection | the function for merging operation result into the collection |
initialCollection | Collection | initial collection which would be the first argument passed into the first invocation of the collect function |
initialParameter | Partial<Param> | the parameter for the first operation |
Promise<Collection>
Promise of collection of all the results returned by the operation function
▸ Static synchronised<T>(lock, operation): Promise<T>
This is just another spelling of PromiseUtils.synchronized.
| Name |
|---|
T |
| Name | Type | Description |
|---|---|---|
lock | unknown | the object (could be a string, a number, or this in a class) that is used to apply the lock |
operation | (previousState: undefined | PromiseState, previousSettledState: undefined | PromiseState, previousResult: any) => Promise<T> | function for doing the computation and returning a Promise |
Promise<T>
the result of the operation function
▸ Static synchronized<T>(lock, operation): Promise<T>
Equivalent of synchronized in Java.
In any situation there's no concurrent execution of any operation function associated with the same lock.
The operation function has access to the state (when synchronized is called), settledState (when the operation function is called),
and result (could be the fulfilled result or the rejected reason) of the previous operation.
In case there is no previous invocation, state, settledState and result would all be undefined.
| Name |
|---|
T |
| Name | Type | Description |
|---|---|---|
lock | unknown | the object (could be a string, a number, or this in a class) that is used to apply the lock |
operation | (previousState: undefined | PromiseState, previousSettledState: undefined | PromiseState, previousResult: any) => Promise<T> | function for doing the computation and returning a Promise |
Promise<T>
the result of the operation function
▸ Static timeoutReject<T, R>(operation, ms, rejectReason): Promise<T>
Apply timeout to a Promise. In case timeout happens, reject with the reason specified. If timeout does not happen, the resolved result or rejection reason of the original Promise would be the outcome of the Promise returned from this function. If timeout does not happen and the 'rejectReason' parameter is a function, the function won't be called. The 'operation' parameter's rejection would not be handled by this function, you may want to handle it outside of this function, just for avoiding warnings like "(node:4330) PromiseRejectionHandledWarning: Promise rejection was handled asynchronously".
| Name | Type |
|---|---|
T | never |
R | any |
| Name | Type | Description |
|---|---|---|
operation | Promise<T> | the original Promise for which timeout would be applied |
ms | number | number of milliseconds for the timeout |
rejectReason | R | PromiseLike<R> | () => R | PromiseLike<R> | the reason of the rejection in case timeout happens, or a function that supplies the reason. |
Promise<T>
the new Promise that rejects with the specified reason in case timeout happens
▸ Static timeoutResolve<T>(operation, ms, result?): Promise<T>
Apply timeout to a Promise. In case timeout happens, resolve to the result specified. If timeout does not happen, the resolved result or rejection reason of the original Promise would be the outcome of the Promise returned from this function. If timeout does not happen and the 'result' parameter is a function, the function won't be called. The 'operation' parameter's rejection would not be handled by this function, you may want to handle it outside of this function, just for avoiding warnings like "(node:4330) PromiseRejectionHandledWarning: Promise rejection was handled asynchronously".
| Name |
|---|
T |
| Name | Type | Description |
|---|---|---|
operation | Promise<T> | the original Promise for which timeout would be applied |
ms | number | number of milliseconds for the timeout |
result? | T | PromiseLike<T> | () => T | PromiseLike<T> | the result to be resolved in case timeout happens, or a function that supplies the reuslt. |
Promise<T>
the new Promise that resolves to the specified result in case timeout happens
▸ Static withRetry<Result, TError>(operation, backoff, shouldRetry?): Promise<Result>
Do an operation repeatedly until a criteria is met.
example
const result = await PromiseUtils.withRetry(() => doSomething(), [100, 200, 300, 500, 800, 1000]);
const result2 = await PromiseUtils.withRetry(() => doSomething(), PromiseUtils.FIBONACCI_SEQUENCE, err => err.statusCode === 429);
| Name | Type | Description |
|---|---|---|
Result | Result | type of the operation result |
TError | any | type of the possible error that could be generated by the operation |
| Name | Type | Description |
|---|---|---|
operation | (attempt: number, previousResult: undefined | Result, previousError: undefined | TError) => Promise<Result> | a function that outputs a Promise result, normally the operation does not use its arguments |
backoff | number[] | (attempt: number, previousResult: undefined | Result, previousError: undefined | TError) => undefined | number | Array of retry backoff periods (unit: milliseconds) or function for calculating them. If retry is desired, before making next call to the operation the desired backoff period would be waited. If the array runs out of elements or the function returns undefined or either the array or the function returns a negative number, there would be no further call to the operation. The attempt argument passed into backoff function starts from 2 because only retries need to backoff, so the first retry is the second attempt. |
shouldRetry | (previousError: undefined | TError, previousResult: undefined | Result, attempt: number) => boolean | Predicate function for deciding whether another call to the operation should happen. If this argument is not defined, retry would happen whenever the operation rejects with an error. shouldRetry would be evaluated before backoff. The attempt argument passed into shouldRetry function starts from 1. |
Promise<Result>
Promise of the operation result potentially with retries already applied
@handy-common-utils/promise-utils / PromiseState
The state of a Promise can only be on of: Pending, Fulfilled, and Rejected.
• Fulfilled = "Fulfilled"
• Pending = "Pending"
• Rejected = "Rejected"
FAQs
Promise related utilities
The npm package @handy-common-utils/promise-utils receives a total of 0 weekly downloads. As such, @handy-common-utils/promise-utils popularity was classified as not popular.
We found that @handy-common-utils/promise-utils demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 0 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.