load-grunt-config

load-grunt-config

load-grunt-config is a Grunt library that allows you to break up your Gruntfile config by task. For most small projects a single Gruntfile.js is perfect. But as a project grows, the Gruntfile.js can quickly become unmanagable; this is where load-grunt-config comes in handy. It was heavily inspired by Thomas Boyt's "More Maintainable Gruntfiles".

Features

• Each task has its own config file. Example: jshint.js, mocha.js, etc.
• Auto load all grunt plugins. Uses load-grunt-tasks. (Optionally it can use jit-grunt)
• Auto expose package.json (<%= package.name %>).
• Support for YAML files.
• Support for CSON files.
• Support for coffeescript files.
• Support for returning a function.
• Easily register task aliases with aliases.(js|json|yaml|coffee).
• Config overrides
• Config grouping

Installation

npm install -D load-grunt-config

Example

Basic Gruntfile.js

module.exports = function(grunt) {

	require('load-grunt-config')(grunt);

};

Gruntfile.js with options

module.exports = function(grunt) {
	var path = require('path');

	require('load-grunt-config')(grunt, {
		// path to task.js files, defaults to grunt dir
		configPath: path.join(process.cwd(), 'grunt'),
		
		// path to project package.json file
		packageJsonPath: path.join(process.cwd(), 'package.json'),

		// auto grunt.initConfig
		init: true,

		// data passed into config.  Can use with <%= test %>
		data: {
			test: false
		},

		// use different function to merge config files
		mergeFunction: require('recursive-merge'),

		// can optionally pass options to load-grunt-tasks.
		// If you set to false, it will disable auto loading tasks.
		loadGruntTasks: {
		
			pattern: 'grunt-*',
			config: require('./package.json'),
			scope: 'devDependencies'
		},

		//can post process config object before it gets passed to grunt
		postProcess: function(config) {},

		//allows to manipulate the config object before it gets merged with the data object
		preMerge: function(config, data) {}
	});

};

Optionally you can use jit-grunt instead of load-grunt-tasks

module.exports = function(grunt) {

	require('load-grunt-config')(grunt, {
		// ...
		jitGrunt: {
		    // here you can pass options to jit-grunt (or just jitGrunt: true)
		    staticMappings: {
		        // here you can specify static mappings, for example:
		        sprite: 'grunt-spritesmith',
                hello: 'custom/say-hello.js'
		    }
		}
	});

};

Note: if you have problems with auto loading of some tasks please check jit-grunt#static-mappings

Grunt tasks files

Here's what the files in your grunt/ folder could look like. You can use either .js, .json, .yaml, .cson, or .coffee - whatever you prefer and you can mix and match as you see fit.

Example js file returning an object - grunt/watch.js

module.exports = {
  all: {
    files: [
      '<%= jshint.all %>',
      'grunt/*.yaml'
    ],
    tasks: [
      'default'
    ]
  }
};

Example js file returning a function - grunt/jshint.js

module.exports = function (grunt, options) {
  return {
    all: [
      'Gruntfile.js',
      'grunt/*.js',
      'lib/*.js',
      'test/*.js',
      options.someFile
    ]
  };
};

Example json file - grunt/clean.json

{
  "all": [
    "<%= project.dest %>",
    "target/*.js"
  ]
}

Example yaml file - grunt/notify.yaml

default:
  options:
    message: 'Default finished'

Example coffee file - grunt/task.coffee

module.exports =
  options:
    bare: true

Aliases

If your grunt/ folder contains an aliases.(js|.json|yaml|cson|coffee) file, load-grunt-config will use that to define your tasks aliases (like grunt.registerTask('default', ['jshint']);).

The following examples show the same aliasses definition written in various formats

Example yaml file - grunt/aliases.yaml

default: []

lint:
  description: 'Helps to make our code better'
  tasks:
    - 'jshint'
    - 'csslint'

build:
  - 'lint'
  - 'mocha'
  - 'notify'

Example json file - grunt/aliases.json

{
  "default": [],
  "lint": [
    "jshint",
    "csslint"
  ],
  "build": [
    "lint",
    "mocha",
    "notify"
  ]
}

Example JavaScript file returning an object - grunt/aliases.js

module.exports = {
  'default': [],
  'lint': [
    'jshint',
    'csslint'
  ],
  'build': [
    'lint',
    'mocha',
    'notify'
  ]
};

Example JavaScript file returning a function grunt/aliases.js Useful if there is need to compute something before return.

module.exports = function (grunt, options) {
  // computation...
  return {
    'default': [],
    'lint': [
      'jshint',
      'csslint'
    ],
    'build': [
      'lint',
      'mocha',
      'notify'
    ]
  };
};

Example coffee file grunt/aliases.coffee

module.exports =
  default: []
  lint: [
    'jshint'
    'csslint'
  ]
  build: [
    'lint'
    'mocha'
    'notify'
  ]

You can specify a task description - example JavaScript file grunt/aliases.js

module.exports = {
  'lint': {
    description: 'Lint css and js',
    tasks: [
      'jshint',
      'csslint'
    ]
  }
};

Custom Config

There are certain scenarios where you might have a base config for your team, and you want to be able to override some of the config based on your personal setup. You can do that with the overridePath property. In this case, the library will merge the two, with the override path taking priority. For example:

module.exports = function(grunt) {
  var path = require('path');
  
  require('load-grunt-config')(grunt, {
    configPath: path.join(process.cwd(), 'vendor'),
    overridePath: path.join(process.cwd(), 'config-'+process.env.USER)
  });

};

configPath and overridePath accept single string as well as array of strings. It means that you can compose config using multiple folders. For example:

module.exports = function(grunt) {
  var path = require('path');
  
  require('load-grunt-config')(grunt, {
    configPath: [
      path.join(process.cwd(), 'vendor'),
      path.join(process.cwd(), 'base-target')
    ],
    overridePath: [
      path.join(process.cwd(), 'variant-1'),
      path.join(process.cwd(), 'variant-n')
    ]
  });

};

Config Grouping

load-grunt-config also supports grouping tasks. This is handy when you want to group all of your script or css tasks together. To do that, just add the suffix -tasks to your config filename and load-grunt-config will treat the filename as the task target and the top level keys as the task names.

Here's an example

Filename: /config/scripts-tasks.yaml

jshint:
  files:
    - '*.js'
jshint__test:
  files:
    - 'test/*.js'
watch:
  files:
    - '*.js'
  tasks:
    - 'scripts'

This would be the equivalent in your Gruntfile.js:

{
  jshint: {
    scripts: {
      files: [
        '*.js'
      ]
    },
    scripts_test: {
      files: [
        'test/*.js'
      ]
    }
  },
  watch: {
    scripts: {
      files: [
        '*.js'
      ],
      tasks: [
        'scripts'
      ]
    }
  }
}

Debugging

If you pass the parameter --config-debug, load-grunt-config will output the whole object it will pass to Grunt, which can be useful for debugging purposes or when asking for help.

Note that this won't run grunt at all and no tasks would be run, nor loaded.