Your code should always be clean when doing npm run lint. It runs eslint.
test
├── integration # Integration tests (only ran on release)
│ ├── browser # tests ran in browser environment
│ ├── common # tests ran in both node and browser environment
│ ├── node # tests only ran in the node environment
│ └── utils # files containing configuration and mocks
└── unit # Unit tests (run on Travis)
├── browser # files needed by tests
├── common # files needed by tests
├── node # files needed by tests
└── utils # files needed by testsWe use mocha.
To run individual tests:
# Whole file
mocha test/unit/common/client_test.js
# Whole dir
mocha test/unit/common/*_test.js
# All unit tests
npm test
They ensure that the client is correctly talking to the stream API which is necessary for most checks.
They are great because they mimic a real setup where someone would perform an API request to the Stream API. Using mocks or pre-saved responses often hides corner-cases and are the source of lots of issues.
We like to see that the metrics being generated are actually covered by tests. To generate the coverage report you can run npm run coverage, this will invoke the nodejs istanbul coverage reporter and store its output in /coverage.
Coverage is also tracked by coveralls.io. Note, however, that Travis does not enforce coverage. It will not fail the test run if you don't have 100% coverage.
Its configuration is stored in .travis.yml.
It's running unit tests inside the Node environment.
Its configuration is stored in karma.ci.config.js.
It's using the karma test runner to execute browser unit tests inside several different browsers on multiple platforms. We use a karma plugin to enable saucelabs integration for karma.
To add a new browser to the test run edit the karma configuration file.
Please read first the integration test description.