atlas-parallel
v1.0.1
Published
Run async functions in parallel with an onDone callback.
Downloads
16
Maintainers
Readme
atlas-parallel
Run async functions in parallel with an onDone callback.
install
npm install --save atlas-parallel
why
For a lot of applications, all I need is a simple parallel function runner and I don't wanna import something like async or q, because I rarely use any of the other functions. This package is literally just a parallel runner and nothing else. In other words, it will run N async functions in (roughly) the time it takes to run the one with the highest latency.
examples
Usage is pretty simple -- just pass in a collection of jobs (which each take a done
callback) and an allDone
callback. The allDone
callback is called after each of the done
callbacks have been called.
dictionary of jobs
You can pass in a hash of name -> job pairs and the end result will be a hash of name -> err pairs and/or a hash of name -> result pairs:
const parallel = require("atlas-parallel");
const jobs = {
sendEmail: done => {
email.send("[email protected]", "hello", err => {
err ? done(err) : done(null, "success")
})
},
postPic: done => reddit.post("atlassubbed.png", "mildlyinteresting", err => {
err ? done(err) : done(null, "success")
})
}
parallel(jobs, (errs, res) => {
// all done!
// errs === {sendEmail: null, postPic: null}
// res === {sendEmail: "success", postPic: "success"}
})
array of jobs
If you don't care that much about return values, it is much more concise to use an array of jobs:
...
parallel([
done => email.send("[email protected]", "hello", done),
done => reddit.post("atlassubbed.png", "mildlyinteresting", done)
], errs => {
// all done!
// errs === [] on success
// errs === [err2] if job2 fails
// errs === [err1, err2] if both fail
})
caveats
If you need to keep track of specific errors and return values, use a dictionary of jobs instead of an array. If you forget to call done
on one of your jobs, your allDone
callback will never be called. I haven't implemented any timeout feature -- this could easily be implemented externally and is outside the scope of this package.