Functional Promises

Updated Documentation: fpromises.io

Summary

The Functional Promises library is a Fluent API supporting a specific Function Chaining technique (using composition).

Note: FP doesn't replace or extend Promises. It uses them.

There are many names for this general pattern. Including Collection Pipeline and Promise Chain. The emphasis here is a seamless async/sync Developer Experience.

I call this pattern a Functional River - your data is the water, and your functions describe its path or riverbed.

Advantages

• Only 400 Lines of Source & 3Kb compressed
• Easily handle Sync, Async, Events, Promises, and Callbacks.
• Familiar methods, including Array.prototype.map, [].filter(), [].find(), [].some(), etc.
• Create Monads in JavaScript (so far as they contain side-effects to a sequence of instructions).
• Point-free code is easily achieved (no temp variables).
• Higher code test coverage w/ less repetitive code in tests.
• Use the best features from multiple programming styles: a little imperative, plenty functional, a pinch of OOP, yet still resembling declarative methods!!!

Library Comparison

Total Lines of Code (LoC) calculated using cloc CLI utility.

LoC #'s included because a smaller surface === fewer places bugs can hide.

Library	Main deal	Files	Lines of Code	.min.js kB
Functional Promise v1.8.1	Sync & Async Chains	8	375	12 Kb (3Kb compressed)
Bluebird v3.5.1	Promises Replacement	38	5,188	80 Kb
RxJS v5.5.6	Observables Chaining	458	12,266	150 Kb
IxJS v2.3.4	[Async]Iterable Chaining	521	12,366	145 Kb

Admittedly IxJS/RxJS have a far larger API than FP also some behavior in RxJS/IxJS may never be added. Currently however there is a lot of overlap with FP (plus more planned).

The table above show FP is roughly 1/30th the LOC (lines of code) in R/IxJs. FP's bundle size is about 10% the size of either RxJS/IxJS.

BluebirdJS and FP have roughly the same number of API methods, yet Bluebird has a fair bit more code.

To be clear: Bluebird and RxJS/IxJS are amazing. Their interface/designs has been very influential on FP.

Note: R/IxJS's hyper-modular design also allows for bundles to be lots smaller (though using quite different syntax, either .pipe(...) or ix/iterable/ix/add/...).

Installation

npm install functional-promises

Getting Started

Use one of the following:

const FP = require('functional-promises')
// or:
import FP from 'functional-promises'

Quick Examples

Using .map()

FP.resolve([1, 2, 3, 4, 5])
  .map(x => x * 2)
  .map(x => x * 2)
  .then(results => {
    // results === [4, 8, 12, 16, 20]
  })

Handling Events

Create function chains to handle the case where promises don't fit very naturally.

For example streams & event handlers must (usually) support multiple calls over time.

Here's how FP.chain() and FP.chainEnd()/FP.listen(obj, event) help you handle this like a pro:

const button = document.getElementById('submitBtn')
FP.chain() // start a chain
  .then(({target}) => { // destructure 'target' from the `event`
    target.textContent = 'Clicked!'
  })
  .listen(button, 'click') // end the repeatable chain, started at `.chain()`

API Outline

• Thenable Methods
  • Arrays
    • .map(fn)
    • .filter(fn)
    • .find(fn)
    • .findIndex(fn)
    • .some(fn)
    • .none(fn)
    • .series(fn)
  • Errors
    • .catch(fn)
    • .catch(filter, fn)
  • Conditional
    • .thenIf(fn, ifTrue, ifFalse)
  • Utilities
    • .tap(fn)
    • .delay(msec)
  • Properties
    • .get(keyName)
    • .set(keyName, value)
• Specialty Methods
  • Helpers
    • FP.resolve()
    • FP.all(Object/Array)
    • FP.unpack()
  • Events
    • .listen(obj, ...eventNames)
  • Composition Pipelines
    • FP.chain(options)
    • .chainEnd()
  • Modifiers
    • .quiet() - prevents errors from stopping array methods mid-loop
    • .concurrency(threadLimit) - limits parallel workers for array methods

Development

git clone [e-mail]:functional-promises/functional-promises.git
cd functional-promises
npm install
npm test

Thanks to several influencial projects: RxJS, IxJS, Bluebird, asynquence, FantasyLand, Gulp, HighlandJS, et al.

Special thanks to Kyle Simpson, Eric Elliot, and Sarah Drasner for their work for the OSS community, as well as their advice & encouragement.