Async2.js

Better asynchronous javascript flow control in 98 lines or 2.59KB minified (990 bytes gzipped).

Inspired by async, mini-async, Mocha, Chai, Should.js, and IcedCoffeeScript/TameJs libraries.

Flow Control

• begin / try / new : chainable instantiation; not required but sometimes useful
• beforeAll / before : non-blocking function called once before first task
• beforeEach : non-blocking function called once before each task
• serial / series / blocking / waterfall : blocking function called in order; results optionally waterfalled.
• parallel / nonblocking : non-blocking function called in order
• do / then / auto : optionally blocking function called in order; determined by length of arguments callback expects
• afterEach / between / inbetween : non-blocking function called once after each task
• error / catch / rescue : blocking function called when error occurs
• success / else : non-blocking function called after all tasks have completed, but only if no errors occur
• end / finally / ensure / afterAll / after / complete / done : blocking function called after all tasks have completed
• whilst : provide test, iterator, and callback functions. will iterate until test passes, then execute callback

Quick Examples

First, reflect upon our haiku mantra:

"thoughtful single-chain

order of operations

escape callback hell!"

Then, observe in action:

async('pretend/path/to/file') # accepts initial input passed via waterfall to serial
  .serial(fs.readFile) # block until data is read
  .parallel [ tweet, fbook, gplus ], # push data to 3 services simultaneously
    (err, data) -> # execute once all above is complete
      #done()

For the latest examples, review the easy-to-follow ./test/test.coffee.

Or try it immediately in your browser with codepen.

TODO

• potential node.js madness: each series becomes its own cpu thread, each parallel becomes its own gpu thread.

"GPUs have evolved to the point where many real-world applications are easily implemented on them and run significantly faster than on multi-core systems. Future computing architectures will be hybrid systems with parallel-core GPUs working in tandem with multi-core CPUs.' -- Professor Jack Dongarra, Director of the Innovative Computing Laboratory, The University of Tennessee