Organizing Test Suite
As projects grow, inevitably more and more integration tests are added. This increases build time and slows productivity.
To prevent this, you should run your tests in parallel. WebdriverIO already tests each spec (or feature file in Cucumber) in parallel within a single session. In general, try to test only a single feature per spec file. Try to not have too many or too few tests in one file. (However, there is no golden rule here.)
Once your tests have several spec files, you should start running your tests concurrently. To do so, adjust the maxInstances
property in your config file. WebdriverIO allows you to run your tests with maximum concurrency—meaning that no matter how many files and tests you have, they can all run in parallel. (This is still subject to certain limits, like your computer’s CPU, concurrency restrictions, etc.)
Let's say you have 3 different capabilities (Chrome, Firefox, and Safari) and you have set
maxInstances
to1
. The WDIO test runner will spawn 3 processes. Therefore, if you have 10 spec files and you setmaxInstances
to10
, all spec files will be tested simultaneously, and 30 processes will be spawned.
You can define the maxInstances
property globally to set the attribute for all browsers.
If you run your own WebDriver grid, you may (for example) have more capacity for one browser than another. In that case, you can limit the maxInstances
in your capability object:
// wdio.conf.js
export const config = {
// ...
// set maxInstance for all browser
maxInstances: 10,
// ...
capabilities: [{
browserName: 'firefox'
}, {
// maxInstances can get overwritten per capability. So if you have an in-house WebDriver
// grid with only 5 firefox instance available you can make sure that not more than
// 5 instance gets started at a time.
browserName: 'chrome'
}],
// ...
}
Inherit From Main Config File
If you run your test suite in multiple environments (e.g., dev and integration) it may help to use multiple configuration files to keep things manageable.
Similar to the page object concept, the first thing you’ll need is a main config file. It contains all configurations you share across environments.
Then create another config file for each environment, and supplement the the main config with the environment-specific ones:
// wdio.dev.config.js
import { deepmerge } from 'deepmerge-ts'
import wdioConf from './wdio.conf.js'
// have main config file as default but overwrite environment specific information
export const config = deepmerge(wdioConf.config, {
capabilities: [
// more caps defined here
// ...
],
// run tests on sauce instead locally
user: process.env.SAUCE_USERNAME,
key: process.env.SAUCE_ACCESS_KEY,
services: ['sauce']
}, { clone: false })
// add an additional reporter
config.reporters.push('allure')
Grouping Test Specs In Suites
You can group test specs in suites and run single specific suites instead of all of them.
First, define your suites in your WDIO config:
// wdio.conf.js
export const config = {
// define all tests
specs: ['./test/specs/**/*.spec.js'],
// ...
// define specific suites
suites: {
login: [
'./test/specs/login.success.spec.js',
'./test/specs/login.failure.spec.js'
],
otherFeature: [
// ...
]
},
// ...
}
Now, if you want to only run a single suite, you can pass the suite name as a CLI argument:
wdio wdio.conf.js --suite login
Or, run multiple suites at once:
wdio wdio.conf.js --suite login --suite otherFeature
Grouping Test Specs To Run Sequentially
As described above, there are benefits in running the tests concurrently. However, there are cases where it would be beneficial to group tests together to run sequentially in a single instance. Examples of this are mainly where there is a large setup cost e.g. transpiling code or provisioning cloud instances, but there are also advanced usage models that benefit from this capability.
To group tests to run in a single instance, define them as an array within the specs definition.
"specs": [
[
"./test/specs/test_login.js",
"./test/specs/test_product_order.js",
"./test/specs/test_checkout.js"
],
"./test/specs/test_b*.js",
],
In the example above, the tests 'test_login.js', 'test_product_order.js' and 'test_checkout.js' will be run sequentially in a single instance and each of the "test_b*" tests will run concurrently in individual instances.
It is also possible to group specs defined in suites, so you can now also define suites like this:
"suites": {
end2end: [
[
"./test/specs/test_login.js",
"./test/specs/test_product_order.js",
"./test/specs/test_checkout.js"
]
],
allb: ["./test/specs/test_b*.js"]
},
and in this case all of the tests of the "end2end" suite would be run in a single instance.
When running tests sequentially using a pattern, it will run the spec files in an alphabetical order
"suites": {
end2end: ["./test/specs/test_*.js"]
},
This will run the files matching the pattern above in the following order:
[
"./test/specs/test_checkout.js",
"./test/specs/test_login.js",
"./test/specs/test_product_order.js"
]
Run Selected Tests
In some cases, you may wish to only execute a single test (or subset of tests) of your suites.
With the --spec
parameter, you can specify which suite (Mocha, Jasmine) or feature (Cucumber) should be run. The path is resolved relative from your current working directory.
For example, to run only your login test:
wdio wdio.conf.js --spec ./test/specs/e2e/login.js
Or run multiple specs at once:
wdio wdio.conf.js --spec ./test/specs/signup.js --spec ./test/specs/forgot-password.js
If the --spec
value does not point to a particular spec file, it is instead used to filter the spec filenames defined in your configuration.
To run all specs with the word “dialog” in the spec file names, you could use:
wdio wdio.conf.js --spec dialog
Note that each test file is running in a single test runner process. Since we don't scan files in advance (see the next section for information on piping filenames to wdio
), you can't use (for example) describe.only
at the top of your spec file to instruct Mocha to run only that suite.
This feature will help you to accomplish the same goal.
When the --spec
option is provided, it will override any patterns defined by the config or capability level's specs
parameter.
Exclude Selected Tests
When needed, if you need to exclude particular spec file(s) from a run, you can use the --exclude
parameter (Mocha, Jasmine) or feature (Cucumber).
For example, to exclude your login test from the test run:
wdio wdio.conf.js --exclude ./test/specs/e2e/login.js
Or, exclude multiple spec files:
wdio wdio.conf.js --exclude ./test/specs/signup.js --exclude ./test/specs/forgot-password.js
Or, exclude a spec file when filtering using a suite:
wdio wdio.conf.js --suite login --exclude ./test/specs/e2e/login.js
If the --exclude
value does not point to a particular spec file, it is instead used to filter the spec filenames defined in your configuration.
To exclude all specs with the word “dialog” in the spec file names, you could use:
wdio wdio.conf.js --exclude dialog
When the --exclude
option is provided, it will override any patterns defined by the config or capability level's exclude
parameter.
Run Suites and Test Specs
Run an entire suite along with individual specs.
wdio wdio.conf.js --suite login --spec ./test/specs/signup.js
Run Multiple, Specific Test Specs
It is sometimes necessary—in the context of continuous integration and otherwise—to specify multiple sets of specs to run. WebdriverIO's wdio
command line utility accepts piped-in filenames (from find
, grep
, or others).
Piped-in filenames override the list of globs or filenames specified in the configuration's spec
list.
grep -r -l --include "*.js" "myText" | wdio wdio.conf.js
Note: This will not override the --spec
flag for running a single spec.
Running Specific Tests with MochaOpts
You can also filter which specific suite|describe
and/or it|test
you want to run by passing a mocha specific argument: --mochaOpts.grep
to the wdio CLI.
wdio wdio.conf.js --mochaOpts.grep myText
wdio wdio.conf.js --mochaOpts.grep "Text with spaces"
Note: Mocha will filter the tests after the WDIO test runner creates the instances, so you might see several instances being spawned but not actually executed.
Exclude Specific Tests with MochaOpts
You can also filter which specific suite|describe
and/or it|test
you want to exclude by passing a mocha specific argument: --mochaOpts.invert
to the wdio CLI. --mochaOpts.invert
performs opposite of --mochaOpts.grep
wdio wdio.conf.js --mochaOpts.grep "string|regex" --mochaOpts.invert
wdio wdio.conf.js --spec ./test/specs/e2e/login.js --mochaOpts.grep "string|regex" --mochaOpts.invert
Note: Mocha will filter the tests after the WDIO test runner creates the instances, so you might see several instances being spawned but not actually executed.
Stop testing after failure
With the bail
option, you can tell WebdriverIO to stop testing after any test fails.
This is helpful with large test suites when you already know that your build will break, but you want to avoid the lengthy wait of a full testing run.
The bail
option expects a number, which specifies how many test failures can occur before WebDriver stop the entire testing run. The default is 0
, meaning that it always runs all tests specs it can find.
Please see Options Page for additional information on the bail configuration.
Run options hierarchy
When declaring what specs to run, there is a certain hierarchy defining what pattern will take precedence. Currently, this is how it works, from highest priority to lowest:
CLI
--spec
argument > capabilityspecs
pattern > configspecs
pattern CLI--exclude
argument > configexclude
pattern > capabilityexclude
pattern
If only the config parameter is given, it will be used for all capabilities. However, if defining the pattern at the capability level, it will be used instead of the config pattern. Finally, any spec pattern defined on the command line will override all other patterns given.
Using capability-defined spec patterns
When you define a spec pattern at the capability level, it will override any patterns defined at the config level. This is useful when needing to separate tests based on differentiating device capabilities. In cases like this, it is more useful to use a generic spec pattern at the config level, and more specific patterns at the capability level.
For example, let's say you had two directories, with one for Android tests, and one for iOS tests.
Your config file may define the pattern as such, for non-specific device tests:
{
specs: ['tests/general/**/*.js']
}
but then, you will have different capabilities for your Android and iOS devices, where the patterns could look like such:
{
"platformName": "Android",
"specs": [
"tests/android/**/*.js"
]
}
{
"platformName": "iOS",
"specs": [
"tests/ios/**/*.js"
]
}
If you require both of these capabilities in your config file, then the Android device will only run the tests under the "android" namespace, and the iOS tests will run only tests under the "ios" namespace!
//wdio.conf.js
export const config = {
"specs": [
"tests/general/**/*.js"
],
"capabilities": [
{
platformName: "Android",
specs: ["tests/android/**/*.js"],
//...
},
{
platformName: "iOS",
specs: ["tests/ios/**/*.js"],
//...
},
{
platformName: "Chrome",
//config level specs will be used
}
]
}