Filter low-level Stream instances into stackable, protocol-based streams.


This module exposes the StreamStack interface, which starts off as a node Stream subclass that accepts a "parent" Stream to work with. StreamStack is meant to be subclassed in order to implement the layers of a protocol, or run the parent Stream's data through some kind of filter (i.e. compression).

By default, a StreamStack instance proxies all events downstream (from the parent stream to the child stream), and proxies all functions calls upstream (from the child stream to the parent stream).

Keeping the StreamStack subclass' implementation independent of the parent Stream instance allows for the backend transport to be easily swapped out for flexibility and code re-use. For example, storing netcat results to a file, and using fs.ReadStream as your parent stream, rather than net.Stream, in your test cases.

Since StreamStack inherits from the regular node Stream, all it's prototypal goodies can be used along with your subclass instances. This makes it extremely easy for you to call Stream#pipe(writable), in order to utilize node's data transfer philosophies.

Here's a simple, kinda silly example:

var util = require('util');
var StreamStack = require('stream-stack').StreamStack;
// The first argument is the parent stream 
function DoubleWrite(stream) {, stream);
util.inherits(DoubleWrite, StreamStack);
// Overwrite the default `write()` function to call 
// write() on the parent stream twice! 
DoubleWrite.prototype.write = function(data) {;;
// How to Use: 
var doubleStdout = new DoubleWrite(process.stdout);
doubleStdout.write("this will be printed twice!\n");

We've defined a DoubleWrite class. It accepts a writable stream, and whenever write() is called on the DoubleWrite instance, then in return write() get called twice on the parent stream. In this example, our writable stream, process.stdout, will get the string printed to it twice.

Check out the Wiki page to see the list of Known Subclasses.