Grunt task for node-browserify.
grunt-browserify is looking for a new maintainer. If you are interested in taking over, please message github.com/jmreidy or github.com/tleunen. Thanks!
This plugin requires Grunt ~0.4.0.
Install this grunt plugin with:
npm install grunt-browserify --save-devThen add this line to your project's grunt.js Gruntfile:
grunt.loadNpmTasks('grunt-browserify');An important note for those running the latest release of grunt-browserify:
the newest version (>3.0) incorporates breaking changes from Browserify
which REMOVED BUNDLE OPTIONS. All options to browserify must now be passed
in the browserifyOptions hash.
Most simply, Browserify is a tool for taking your CommonJS-style Javascript code and packaging it for use in the browser. Grunt-Browserify provides the glue to better integrate Browserify into your Grunt-based development workflow.
For JavaScripters unfamiliar with CJS-style code and the Node ecosystem, moving to Browserify can be a bit confusing. Writing your client-side code as CJS modules allows for smaller, easier to understand files that perform one task well. These modules, because of their simplicity, will be significantly easier to use across projects. CJS modules also help to expose the dependency graph inherent in your code, allowing you to write cleaner, more-maintainable modules. As Alex MacCaw writes:
CommonJS modules are one of the best solutions to JavaScript dependency management.
CommonJS modules solve JavaScript scope issues by making sure each module is executed in its own namespace. Modules have to explicitly export variables they want to expose to other modules, and explicitly import other modules; in other words, there's no global namespace.
(A note to AMD fans that the benefits above are not unique to the CJS style of writing JavaScript modules, but the ease-of-interoperability with Node.JS code is a plus of CJS.)
As you begin to write your client-side code in small, reusable modules, you start to have a lot more files to manage. At the same time, you need to integrate these files with other client-side libraries, some of which do not play particularly nicely with a CJS module system. The simplicity provided by CJS modules can be lost as build complexity is increased and Browserify compilation time gets out of control.
Run this task with the grunt browserify command. As with other Grunt plugins, the src and dest properties are most important: src will use the Grunt glob pattern to specify files for inclusion in the browserified package, and dest will specify the outfile for the compiled module.
The current version of grunt-browserify sticks as close to the core browserify API as possible. Additional functionality can be added via the rich ecosystem of browserify transforms and plugins.
The following task options are supported:
Type: Object{alias:path}
Browserify can alias files or modules to a certain name. For example, require('./foo') can be aliased to be used as require('foo').
options: {
alias: {
'foo': './foo.js'
}
}The alias option is just a shortcut to require a file and expose a different name for it. You could do exactly the same thing using require instead of alias. It's equivalent to require: [ ['./foo.js', {expose: 'foo'} ] ]
If you need alias mappings, you can use @joeybaker's remapify plugin, as demonstrated in the code below:
options: {
plugin: [
[
'remapify', [{
src: './client/views/**/*.js', // glob for the files to remap
expose: 'views', // this will expose `__dirname + /client/views/home.js` as `views/home.js`
cwd: __dirname // defaults to process.cwd()
}
]
]
]
}Type: String
Default: empty string
The string will be prepended to the output. Template strings (e.g. <%= config.value %> will be expanded automatically.
Type: [String] or [String:String] or [[String, Object]]
Specifies files to be required in the browserify bundle. String filenames are parsed into their full paths with path.resolve. Aliases can be provided by using the filePathString:aliasName format.
Each require can also be provided with an options hash; in this case, the require should be specified as an array of [filePathString, optionsHash].
Type: [String]
Specifies files to be ignored in the browserify bundle.
String filenames are parsed into their full paths with path.resolve.
Globbing patterns are supported.
Type: [String]
Specifies files or modules to be excluded in the browserify bundle. Globbing patterns are supported; globbed filenames are parsed into their full paths.
Type: [String] or Object{alias:path}.
Specifies id strings which will be loaded from a previously loaded, “common” bundle. That is to say, files in the bundle that require the target String will assume that the target is provided externally.
The secondary form of this option
follows the format of alias above, and will externalise the ids specified in
the alias object. This second form allows for the declaration of a single alias
object which can be supplied to one bundle's alias option and another option's
external option.
In either case, globbing patterns are supported.
Type: [String || Function] or [[String || Function, Object]]
Specifies a pipeline of functions (or modules) through which the browserified bundle will be run. The transform can either be a literal function, or a string referring to a NPM module. The browserify docs themselves explain transform well, but below is an example of transform used with grunt-browserify to automatically compile coffeescript files for use in a bundle:
browserify: {
dist: {
files: {
'build/module.js': ['client/scripts/**/*.js', 'client/scripts/**/*.coffee']
},
options: {
transform: ['coffeeify']
}
}
}Transforms can also be provided with an options hash; in this case, the transform should be specified as an array of [transformStringOrFn, optionsHash].
Note for browserify-shim, the configuration of this transformation has to be inside package.json. Please see documentation of browserify-shim and our example.
Type: [String || Function]
Register a browserify plugin with the bundle. As with transforms, plugins are identified with either their NPM name (String) or a function literal.
Type: Object
A hash of options that are passed to browserify during instantiation. Task-level browserifyOptions are not merged into target-level options.
If a target overrides task-level browserifyOptions, it overrides all of it.
Browserify Github README
Type: Boolean If true, invoke watchify instead of browserify.
For watchify to work properly, you have to keep the process running. The option keepAlive can help you do that, or you can use another grunt-watch task.
Type: Boolean
If true and if watch above is true, keep the Grunt process alive (simulates grunt-watch functionality).
Type: Object A hash of options that are passed to watchify during instantiation. Watchify Github README
Type: Function (b)
An optional callback function that is invoked once before the bundle runs. This can be used for programatically configuring browserify using it's API.
b is the browserify instance for the bundle.
Type: Function (b)
An optional callback function, that will be called before bundle completion.
b is the browerify instance that will output the bundle.
NB: This callback will be invoked every time the bundle is built so when used with the watch option set to true it will be called multiple times. Do not register transforms in this callback or they will end up being registered multiple times.
Type: Function (err, src, next)
An optional callback function, which will be called after bundle completion and
before writing of the bundle. The err and src arguments are provided
directly from browserify. The next callback should be called with (err, modifiedSrc); the modifiedSrc is what will be written to the output file.
NB: This callback will be invoked every time the bundle is built so when used with the watch option set to true it will be called multiple times.
In lieu of a formal styleguide, take care to maintain the existing coding style. Add unit tests for any new or changed functionality. Lint and test your code.
See the CHANGELOG.
Copyright (c) 2013-2015 grunt-browserify contributors
Licensed under the MIT license.