grunt-karma3.0.2 • Public • Published
Grunt plugin for Karma
This current version uses
karma@^3.0.0. For using older versions see the
old releases of grunt-karma.
From the same directory as your project's Gruntfile and package.json, install karma and grunt-karma with the following commands:
$ npm install karma --save-dev$ npm install grunt-karma --save-dev
Once that's done, add this line to your project's Gruntfile:
Gruntfile.js file, add a section named
any number of configurations for running karma. You can either put your
config in a [karma config file] or leave it all in your Gruntfile (recommended).
Here's an example that points to the config file:
Here's an example that puts the config in the Gruntfile:
You can override any of the config file's settings by putting them directly in the Gruntfile:
karma:unit:configFile: 'karma.conf.js'port: 9999singleRun: truebrowsers: 'PhantomJS'logLevel: 'ERROR'
To change the
logLevel in the grunt config file instead of the karma config, use one of the following strings:
files option can be extended "per-target" in the typical way
Grunt handles files:
karma:options:files: 'lib/**/*.js'unit:files:src: 'test/**/*.js'
When using the "Grunt way" of specifying files, you can also extend the file objects with the options supported by karma:
karma:unit:files:src: 'test/**/*.js' served: truesrc: 'lib/**/*.js' served: true included: false
Config with Grunt Template Strings in
When using template strings in the
files option, the results will flattened. Therefore, if you include a variable that includes an array, the array will be flattened before being passed to Karma.
meta:jsFiles: 'jquery.js''angular.js'karma:options:files: '<%= meta.jsFiles %>''angular-mocks.js''**/*-spec.js'
If you have multiple targets, it may be helpful to share common
configuration settings between them. Grunt-karma supports this by
karma:options:configFile: 'karma.conf.js'port: 9999browsers: 'Chrome' 'Firefox'continuous:singleRun: truebrowsers: 'PhantomJS'dev:reporters: 'dots'
In this example the
dev targets will both use
port specified in the
continuous target will override the browser setting to use
PhantomJS, and also run as a singleRun. The
dev target will simply
change the reporter to dots.
There are three ways to run your tests with karma:
Karma Server with Auto Runs on File Change
autoWatch option to true will instruct karma to start
a server and watch for changes to files, running tests automatically:
karma:unit:configFile: 'karma.conf.js'autoWatch: true
$ grunt karma
Karma Server with Grunt Watch
Many Grunt projects watch several types of files using grunt-contrib-watch.
Config karma like usual (without the autoWatch option), and add
karma:unit:configFile: 'karma.conf.js'background: truesingleRun: false
background option will tell grunt to run karma in a child process
so it doesn't block subsequent grunt tasks.
singleRun: false option will tell grunt to keep the karma server up
after a test run.
watch task to run the karma task with the
:run flag. For example:
watch://run unit tests with karma (server needs to be already running)karma:files: 'app/js/**/*.js' 'test/browser/**/*.js'tasks: 'karma:unit:run' //NOTE the :run flag
In your terminal window run
$ grunt karma:unit:start watch, which starts the
karma server and the watch task. Now when grunt watch detects a change to
one of your watched files, it will run the tests specified in the
target using the already running karma server. This is the preferred method
Keeping a browser window & karma server running during development is
productive, but not a good solution for build processes. For that reason karma
provides a "continuous integration" mode, which will launch the specified
browser(s), run the tests, and close the browser(s). It also supports running
tests in PhantomJS, a headless webkit browser which is great for running tests as part of a build. To run tests in continous integration mode just add the
karma:unit:configFile: 'config/karma.conf.js'//continuous integration mode: run tests once in PhantomJS browser.continuous:configFile: 'config/karma.conf.js'singleRun: truebrowsers: 'PhantomJS'
The build would then run
grunt karma:continuous to start PhantomJS,
run tests, and close PhantomJS.
Using additional client.args
You can pass arbitrary
client.args through the commandline like this:
$ grunt karma:dev watch --grep=mypattern