spawn-task-experiment-2 v0.2.0

Spawn task experiment 2

Tests how fast it is to run gulp tasks in child processes

This is a fork of Spawn task experiment to allow to pass some information into child process. I sent a pullRequest but never got an answer, so I'm posting this so I can use it in npm.

Run task in child process:

var makeItFaster = require('spawn-task-experiment').spawn;

gulp.task('whatever', makeItFaster(function() {
	var gulp = require('gulp');
	var sourcemaps = require('gulp-sourcemaps');
	var concat = require('gulp-concat');
	return gulp.src(['app/index.js', 'app/**/*.js'])
		.pipe(sourcemaps.init())
		.pipe(concat('all.js'))
		.pipe(sourcemaps.write('.'))
		.pipe(gulp.dest('.build/scripts'));
}));

Run task in worker pool:

var makeItFaster = require('spawn-task-experiment').workerPool();

gulp.task('whatever', makeItFaster(function() {
	var gulp = require('gulp');
	var sourcemaps = require('gulp-sourcemaps');
	var concat = require('gulp-concat');
	return gulp.src(['app/index.js', 'app/**/*.js'])
		.pipe(sourcemaps.init())
		.pipe(concat('all.js'))
		.pipe(sourcemaps.write('.'))
		.pipe(gulp.dest('.build/scripts'));
}));

Notes:

• All wrapped functions need to be self-contained
• Using worker pool prevents parent process exit, so it might be usable only when watch is running
• Worker pool starts re-using child processes only after maxConcurrentWorkers limit has been reached (default is cpus().length)