Promise.allComplete

[guinness]

So asynchronous programming is something quite difficult to do in JavaScript, but since we now have Promises and async/await, it's becoming increasingly less complicated. Plus there is no need for "callback hell".

So I will assume you understand what Promise.all() is and why it can sometimes be problematic, as it will not wait for all requests to complete and just fails on the first rejected Promise. The following function will still reject on a failed Promise, but the difference being it will wait for all the Promises to complete before resolving or rejecting. If no failure occurred, then the resolved value is the same as Promise.all(), an array of resolved values; otherwise, it returns a completed object (see below for details). Also note that the array might contain empty slots, this is so it's easier to debug which Promise failed in the array, as they're inserted in the same index slot

expandcollapse popup

Promise.allComplete = (iterable) => {
    if (!Array.isArray(iterable)) {
        throw new TypeError('Invalid argument, expected "iterable" to be an array');
    }
    
    const completed = {
        resolved: [],
        rejected: [],
    };

    const wrapResolutionOrRejection = (type, index) => valueOrReason => (completed[type][index] = valueOrReason);
    const wrappedIterable = iterable.map((value, index) =>
        Promise.resolve(value)
            // The rejected wrapper function could be put in the catch, but it's wasteful for our purposes
            .then(wrapResolutionOrRejection('resolved', index), wrapResolutionOrRejection('rejected', index))
    )

    return Promise.all(wrappedIterable)
        .then(() => completed.rejected.length === 0 ? Promise.resolve(completed.resolved) : Promise.reject(completed));
};

// Example
const requests = [
    createPromise(true, 10),
    createPromise(false, 10),
    createPromise(true, 200),
    createPromise(true, 1000),
];

// Rejects on the first Promise which fails, but if you check in the console, it didn't wait
// for the third Promise to successfully complete, as the console log came after the error log was displayed
// Promise.all(requests)
//   .then(values => console.log('Successfully completed', values))
//   .catch(err => console.error('Not successfully completed', err))

// "allComplete" is different, in that it will wait for all the Promises to be completed i.e. resolve and reject,
// then resolve if all Promises were successful or reject if one Promise failed.
// It returns the following data structure:
// {
//  resolved: [...Promises which resolved, and inserted by the associated Promise's index],
//  rejected: [..Promises which rejected, and inserted by the associated Promise's index],
// }
Promise.allComplete(requests)
    .then(completed => console.log('Successfully completed', completed))
    .catch(completed => console.log('Not successfully completed', completed))

function createPromise(isResolved, delay) {
    return new Promise((resolve, reject) => {
        setTimeout(() => {
            console.log(`Promise: "${delay}"`);
            if (isResolved) {
                resolve(delay);
            } else {
                reject(new Error('An unexpected error occurred'));
            }
        }, delay);
    });
}

 

Edited April 22, 2018 by guinness

[TheDcoder]

Do you also have this posted at any other place? Something like GitHub's Gist so I can easy keep track by starring the snippet

[guinness]

Thanks for the interest. Right now this is exclusive only to the AutoIt Forum