The high-level streams library for Node.js and the browser. View the Highland website for more in-depth documentation.
Re-thinking the JavaScript utility belt, Highland manages synchronous and asynchronous code easily, using nothing more than standard JavaScript and Node-like Streams. You may be familiar with Promises, EventEmitters and callbacks, but moving between them is far from seamless. Thankfully, there exists a deeper abstraction which can free our code. By updating the tools we use on Arrays, and applying them to values distributed in time instead of space, we can discard plumbing and focus on the important things. With Highland, you can switch between synchronous and asynchronous data sources at will, without having to re-write your code. Time to dive in!
Made by @caolan, with help and patience from friends :)
This branch tracks the ongoing development of version 3.0, which will feature a rewritten Highland core implementation, extensibility support, limited stream lifecycle, and some breaking changes to certain transforms. See #179 and the 3.x label for more details. New features will only be added to this branch. However, until 3.0 is released, we will still be doing bug fixes for the 2.x releases. See the 2.x branch for those files.
Currently, the code is in a semi-stable state. The only major missing feature
is onDestroy
for higher-level
transforms. To try out the new
goodness, install the next
tag from NPM.
npm install --save highland@next
Highland implements the Fantasy Land Alternative, Filterable, and Monad specifications.
Highland supports
Transducers via the
transduce
transform.
Usage as a Node.js module
var _ = require('highland');
Converting to/from Highland Streams
_([1,2,3,4]).toArray(function (xs) {
// xs is [1,2,3,4]
});
Mapping over a Stream
var doubled = _([1,2,3,4]).map(function (x) {
return x * 2;
});
Reading files in parallel (4 at once)
var data = _(filenames).map(readFile).parallel(4);
Handling errors
data.errors(function (err, rethrow) {
// handle or rethrow error
});
Piping to a Node Stream
data.pipe(output);
Piping in data from Node Streams
var output = fs.createWriteStream('output');
var docs = db.createReadStream();
// wrap a node stream and pipe to file
_(docs).filter(isBlogpost).pipe(output);
// or, pipe in a node stream directly:
var through = _.pipeline(_.filter(isBlogpost));
docs.pipe(through).pipe(output);
Handling events
var clicks = _('click', btn).map(1);
var counter = clicks.scan(0, _.add);
counter.each(function (n) {
$('#count').text(n);
});
Learn more at https://caolan.github.io/highland/