Frameworks

The wdio runner currently supports Mocha, Jasmine (v2.0) and Cucumber. These BDD frameworks are all based on Node.js and need to be installed via NPM. Note that the modules need to be installed at the same place WebdriverIO is installed. If you installed WebdriverIO globally make sure you have the test frameworks installed globally as well.

Within your spec files or step definition you can access the webdriver instance using the global variable browser. You don’t need to initiate or end the Selenium session. This is taken care of by the wdio testrunner.

Using Mocha

If you like to use Mocha you should additionally install an assertion library that supports promises, e.g. Chai As Promise. Initialise that library in the before hook in your configuration file:

1
2
3
4
5
6
7
8
before: function() {
var chai = require('chai');
var chaiAsPromised = require('chai-as-promised');
 
chai.use(chaiAsPromised);
expect = chai.expect;
chai.Should();
}

Once that is done you can write beautiful assertions like:

1
2
3
4
5
6
7
describe('my awesome website', function() {
it('should do some chai assertions', function() {
return browser
.url('http://webdriver.io')
.getTitle().should.eventually.be.equal('WebdriverIO');
});
});

WebdriverIO supports Mochas BDD (default) and TDD interface. If you like to write your specs in TDD language set the ui property in your mochaOpts config to tdd, now your test files should get written like:

1
2
3
4
5
6
7
suite('my awesome website', function() {
test('should do some chai assertions', function() {
return browser
.url('http://webdriver.io')
.getTitle().should.eventually.be.equal('WebdriverIO');
});
});

Make sure to checkout the transfer promises section to enable multiple consecutive assertions. If you want to define specific Mocha settings you can do that by adding mochaOpts to your configuration file.

If you get messages like timeout of 10000ms exceeded. Ensure the done() callback is being called in this test. then you can increasing the timeout settings in mochaOpts.

There are three different approaches how to deal with asynchronicity in your it blocks:

The Good Old Callback Way
1
2
3
4
5
6
7
8
9
10
describe('my feature', function() {
it('will do something', function(done) {
browser
.url('http://google.com')
.getTitle().then(function(title) {
console.log(title);
})
.call(done);
});
});
Using Promises
1
2
3
4
5
6
7
8
9
describe('my feature', function() {
it('will do something', function() {
return browser
.url('http://google.com')
.getTitle().then(function(title) {
console.log(title);
});
});
});
The New Fancy Generator way

Note that you shouldn’t forget the asterisks after function.

1
2
3
4
5
6
7
describe('my feature', function() {
it('will do something', function* () {
yield browser.url('http://google.com');
var title = yield browser.getTitle()
console.log(title);
});
});

Using Jasmine

Jasmine already provides assertion methods you can use with WebdriverIO. Also the way how you deal with asynchronicity is the same like using Mocha. The only difference is that you can change Jasmines defaultTimeoutInterval by settings it in the jasmineNodeOpts in your configuration file.

Intercept Assertion

The Jasmine framework allows it to intercept each assertion in order to log the state of the application or website depending on the result. For example it is pretty handy to take a screenshot everytime an assertion fails. In your jasmineNodeOpts you can add a property called expectationResultHandler that takes a function to execute. The function parameter give you information about the result of the assertion. The following example demonstrate how to take a screenshot if an assertion fails:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
jasmineNodeOpts: {
defaultTimeoutInterval: 10000,
expectationResultHandler: function(passed, assertion) {
 
/**
* only take screenshot if assertion failed
*/
if(passed) {
return;
}
 
var title = assertion.message.replace(/\s/g, '-');
browser.saveScreenshot(('assertionError_' + title + '.png'));
 
}
},

Please note that you can’t stop the test execution to do something async. It might happen that the command takes too much time and the website state has changed. Though usually after 2 another commands the screenshot got taken which gives you still valueable information about the error.

Using Cucumber

If you want to use Cucumber set the framework property to cucumber, either by adding framework: 'cucumber' to the config file or by adding -f cucumber to the command line. Please make sure you have cucumber v0.7.0 installed (globally or as project dependency).

Options for Cucumber such as ‘format’ can be given in the config file with cucumberOpts:

1
2
3
cucumberOpts: {
format: "summary"
}