npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

pipelayer

v0.1.0

Published

Fluently compose pipelines using arbitrary plugins

Downloads

4

Readme

pipelayer

Pipelayer is a tool for "virally" adding plugins to node stream objects, such that they pass on the same plugins to any stream they're .pipe()d to, recursively. This is especially useful for working with gulp, e.g.:

// Use plugins without .pipe()

var gulp      = require('gulp'),
    pipelayer = require('pipelayer').withPlugins({
        src: gulp.src,
        dest: gulp.dest,
        coffee: require('gulp-coffee'),
        uglify: require('gulp-uglify'),
    });

gulp.task("someTask", function() {
  return pipelayer
    .src(['src/*.coffee'])  // No need to .pipe() between plugins!
    .coffee({bare: true})   // Each new stream gets extended with
    .uglify()               // all of the same plugins as the last,
    .dest('dist');          // but is otherwise a normal stream.
});

Defining Plugins

Pipelayer's .withPlugins(obj, names?) static method returns a new, customized version of the pipelayer function, with the specified additional plugins found as own-enumerable properties on obj. (Unless an array of names is given, in which case the named properties will be used as plugins, regardless of whether they are own-properties or enumerable.)

The returned function will have the named properties or methods of obj available as static properties on itself, and will augment any streams passed to it with the same plugins. It will also have a .withPlugins() method, that can be used to create extended versions of itself, recursively.

So, we could have made the previous example even simpler, by telling pipelayer what properties to grab from gulp, and then using auto-plug to load the gulp plugins, e.g.:

var gulp      = require('gulp'),
    pipelayer = require('pipelayer')
        .withPlugins(gulp, ['src', 'dest'])
        .withPlugins(require('auto-plug')('gulp'));

The above code will make src(), dest(), and any modules named gulp-* listed in package.json available as methods on pipelayer, and on any streams it creates or pipes to.

Plugins are "late bound", meaning that the added properties or methods aren't retrieved from obj until the first time they're actually used. (This is especially handy when using auto-plug, which by default returns an object whose properties will lazily require() a relevant plugin module.)

The values of obj's named properties can be any Javascript value, but methods are handled specially:

  • If a plugin method returns a stream, the returned stream is augmented with the same plugins as the current stream or pipelayer function.

  • If the returned stream is writable and the plugin was invoked on a stream, then the returned stream will be .pipe()d to before it's augmented and returned.

All other values just become regular properties or methods of the stream (and of the new pipelayer function).

Composing Pipelines

Normally, when you .pipe() node streams together, you end up with the last stream piped, which means you can't compose transform streams ahead of time: you have to assemble everything at once, which makes it hard to build higher-level transforms out of existing ones.

But sometimes -- especially when working with gulp -- it's helpful to be able to stack a bunch of transform streams together, and use the whole thing as if it were one giant transform stream. Pipelayer lets you do that, just by wrapping the first stream in the pipeline, and ending it with an empty .pipe() call:

// Compose a pipeline from some transform streams

var pipelayer = require('pipelayer'),

    aPipeline = pipelayer(aTransform)   // start the chain w/pipelayer()
                .pipe(anotherTransform)
                .pipe(yetAnotherTransform)
                .pipe();                // and end with empty .pipe()

// Now we can use the combined stream as if it were one stream all along
someReadable.pipe(aPipeline).pipe(someWritable)

Pipelayer does this by augmenting its argument's .pipe() method to carry forward the original start of the pipeline, so that the final .pipe() call can create a combined duplex stream. Each subsequently-piped stream is augmented in the same way, and the resulting stream will emit errors for any errors emitted by the underlying streams.

(To maximize the reusability of the created stream, however, the combined stream will not have any plugins added to it. This lets you use whatever plugins you want inside a library to compose a stream, without those plugins becoming visible to users of the resulting stream.)

This pipeline composition aspect of Pipelayer is completely orthogonal to the plugin-adding aspect. If the pipelayer function used to start the chain has plugins, then each piped stream (but not the combined stream) will have the same plugins. If it doesn't have any plugins, then only the .pipe() method of the streams will be altered. And you can use an empty .pipe() call to create a combined stream, even if the chain was composed entirely from plugin calls, e.g.:

// Compose a pipeline from plugin calls

var uglyCoffee = require('pipelayer')
                    .withPlugins(require('auto-plug')('gulp'))
                    .coffee({bare: true})
                    .uglify()
                    .pipe();    // just end the composition with .pipe()

// `uglyCoffee` is now a plain duplex stream with no plugins or special
// behavior, except that calling `.pipe()` on it with no arguments will
// return the same `uglyCoffee` stream.  As such, we can use it in
// another pipeline, where it will pick up *that* pipeline's plugins:

require('pipelayer')
    .withPlugins(gulp, ['src', 'dest'])
    .src("*.coffee")
    .pipe(uglyCoffee)   
    .dest("dist");

Plugin Precedence and Overriding

Pipelayer does not overwrite existing stream properties, whether own-properties or inherited. This means that if you pipe into the same stream twice, either with the same or different plugin sets, the first set of plugins applied will take precedence, with only the non-overlapping plugins from the second set being available.

If you then pipe from such a stream, the streams you pipe to will gain the same set of plugins: i.e., all the first plugins plus the non-overlapping later plugins -- minus, of course, any properties that already exist on the target stream.

Most programs, of course -- at least most of the time -- only create linear pipe chains, or are only using one set of plugins. In the event that you need to cleanly distinguish between plugin sets, or to change the precedence of applied plugins, bear in mind that you can always use an empty .pipe() call to create a fresh, plugin-free combined stream, before adding it to a pipeline with the plugins you want it to have.