gulp-cdnizer

cdnizer plugin for gulp
This plugin will replace references in HTML and other files with CDN locations. It's flexible without being overly complicated, and handles private and public CDNs. It can use your bower installation to determine file versions.
It also provides optional fallback scripts for failed file loading. By default it can only handle failed JavaScript files, but it shouldn't be too difficult to provide a better script.
WARNING
This plugin does not check incoming files. Do not run it on files that you do not want modified.
Usage
First, install gulp-cdnizer
as a development dependency:
npm install --save-dev gulp-cdnizer
Then, add it to your gulpfile.js
:
var cdnizer = require("gulp-cdnizer");
gulp.src("./src/index.html")
.pipe(cdnizer({
defaultBase: "//my.cdn.host/base",
allowRev: true,
allowMin: true,
files: [
'js/app.js',
{
file: 'vendor/angular/angular.js',
package: 'angular',
test: 'angular',
cdn: '//ajax.googleapis.com/ajax/libs/angularjs/${ major }.${ minor }.${ patch }/angular.min.js'
},
{
file: 'vendor/firebase/firebase.js',
test: 'Firebase',
cdn: '//cdn.firebase.com/v0/firebase.js'
}
]
}))
.pipe(gulp.dest("./dist"));
Alternatively, you can just pass in the files array if you don't need to provide any options, and only have custom files:
gulp.src("./src/index.html")
.pipe(cdnizer([
{
file: 'vendor/angular/angular.js',
package: 'angular',
test: 'angular',
cdn: '//cdnjs.cloudflare.com/ajax/libs/angularjs/${ major }.${ minor }.${ patch }/angular.min.js'
},
{
file: 'vendor/firebase/firebase.js',
test: 'Firebase',
cdn: '//cdn.firebase.com/v0/firebase.js'
}
]))
.pipe(gulp.dest("./dist"));
You can also use globs to define groups of file, and dynamic filename properties:
gulp.src("./src/index.html")
.pipe(cdnizer([{
file: 'vendor/angular/*.js',
package: 'angular',
test: 'angular',
cdn: '//ajax.googleapis.com/ajax/libs/angularjs/${ major }.${ minor }.${ patch }/${ filenameMin }'
}]))
.pipe(gulp.dest("./dist"));
Works great on url()
s in CSS files, too:
gulp.src("./src/css/style.css")
.pipe(cdnizer({
defaultCDNBase: '//my.cdn.url/',
relativeRoot: 'css',
files: ['**/*.{gif,png,jpg,jpeg}']
}))
.pipe(gulp.dest("./dist/css/"));
API
cdnizer(options|files)
Note: If options is an array, it is assumed to be the array of files to process, and the rest of the options are left as their default.
options.defaultCDNBase
Type: String
Default: ''
(disabled)
Used for a default, custom CDN, usually for your own files. This will be used in the defaultCDN property to define the default path for a CDN unless overridden.
options.defaultCDN
Type: String
Default: '${ defaultCDNBase }/${ filepathRel }'
This is the default pattern used for generating CDN links when one isn't provided by a specific file.
options.relativeRoot
Type: String
Default: ''
If you are processing a file that references relative files, or is not rooted to the CDN, you can set relativeRoot
to get correct results.
For example, if you have a CSS file under style/
, and you reference images as ../img/foo.png
, you should set relativeRoot
to style/
. Now if your defaultCDNBase
is //example/
, the image will be resolved to //example/img/foo.png
.
options.allowRev
Type: Boolean
Default: true
Allow for file names with gulp-rev
appended strings, in the form of <file>-XXXXXXXX.<ext>
. If you are using the gulp-rev
plugin, this will automatically match filenames that have a rev string appeneded to them. If you are not using gulp-rev
, then you can disable this by setting allowRev
to false
, which will prevent possible errors in accidentally matching similar file names.
You can always manually configure your globs to include an optional rev string by using the form ?(<rev glob here>)
, such as name?(-????????).ext
for appended revs.
options.allowMin
Type: Boolean
Default: true
Allow for file names that optionally have .min
inserted before the file extension (but after rev, if enabled). This allows you to use the base name for a file, and have cndizer
match the minified name.
options.fallbackScript
Type: String
Default:
<script>
function cdnizerLoad(u) {
document.write('<scr'+'ipt src="'+encodeURIComponent(u)+'"></scr'+'ipt>';
}
</script>
Overwrite the default fallback script. If any of the inputs has a fallback, this script is injected before the first occurence of <link
, <script
, or </head>
in the HTML page. Ignored for files that don't contain <head
.
If you already have a script loader (such as yepnope or Modernizr), you can set this to an empty string and override the fallbackTest
below to use that instead. Of course, this won't help if you are loading those scripts off a CDN and they fail!
options.fallbackTest
Type: String
Default: '<script>if(!(${ test })) cdnizerLoad("${ filepath }");</script>'
Overwrite the default fallback test. Note that all options availble to options.files[].cdn
below are available to this template as well, along with the options.files[].test
string.
options.bowerComponents
Type: String
Default: null
If provided, this is the directory to look for bower components in. If not provided, cdnizer will attempt to look for the .bowerrc
file, and if that is not found or does not specify a directory, it falls back to './bower_components'
.
Once the directory is determined, the script will look for files in <bowerComponents>/bower.json
or <bowerComponents>/.bower.json
to try to determine the version of the installed component.
options.files
Type: Array
Default: (none) required
Array of sources or objects defining sources to cdnize. Each item in the array can be one of two types, a simple glob or object hashmap.
When using a glob, if it matches a source, the defaultCDN
template is applied. Because there is no test
, the script will not have a fallback.
The object hashmap gives you full control, using the following properties:
options.files[].file
Type: String
Default: (none) required
Glob to match against for the file to cdnize. All properties within this object will be applied to all files that match the glob. Globs are matched in a first-come, first-served basis, where only the first matched object hashmap is applied.
options.files[].package
Type: String
Default: (none)
Bower package name for this source or set of sources. By providing the package name, cdnizer will look up the version string of the currently installed bower package, and provide it as a property to the cdn
string. This is done by looking for either the bower.json
or .bower.json
file within your bower components directory.
The benefit of doing it this way is that the version used from the CDN always matches your local copy. It won't be automatically updated to the latest patch version without being tested.
options.files[].cdn
Type: String
Default: options.defaultCDN
Provides a custom CDN string, which can be a simple static string, or contain one or more underscore/lodash template properties to be injected into the string:
version
: if package
was provided, this is the currently installed version from bower.major
: if package
was provided, this is the major version number.minor
: if package
was provided, this is the minor version number.patch
: if package
was provided, this is the patch version number.defaultBase
: the default base provided above. Note that this will never end with a trailing slash.filepath
: the full path of the source, as it exists currently. There is no guarantee about the whether this contains a leading slash or not, so be careful.filepathRel
: the full path of the source, but guaranteed to never have a leading slash.filename
: the name of the file, without any parent directoriesfilenameMin
: the name of the file, without any rev tags (if allowRev
is true), but with a .min
extension added. This won't add a min if there is one already.package
: the bower package name, as provided above.
options.files[].test
Type: String
Default: (none)
If provided, this string will be evaluated within a javascript block. If the result is truthy, then we assume the CDN resource loaded properly. If it isn't, then the original local file will be loaded. This is ignored for files that don't get the fallback script.
See options.fallbackScript
and options.fallbackTest
for more information.
options.matchers
Type: Array
Default: []
Array of custom matchers. Use this to add extra patterns within which you would like to cdn-ize URLs, for example if you have such URLs in data-attributes. The matchers should include regular expressions with three matching groups: 1) Leading characters 2) The actual URL to work on 3) Trailing characters.
Example (matches the data-src
attribute in <img>
tags):
matchers: [ { pattern: /(<img\s.*?data-src=["'])(.+?)(["'].*?>)/gi, fallback: false } ]
You can also specify just a regular expression. In that case, fallback will default to false.
Equivalent example:
matchers: [ pattern: /(<img\s.*?data-src=["'])(.+?)(["'].*?>)/gi ]
Help Support This Project
If you'd like to support this and other OverZealous Creations (Phil DeJarnett) projects, donate via Gittip!

You can learn a little more about me and some of the work I do for open source projects in an article at CDNify.
License
MIT License