Online surveys and reporting
This project is maintained by Nivocer
We use phpunit for unit testing
Tests are definied in /tests (mainly application)
phpunit.xml defines the configuration, including codecoverage
phpunit.travis.xml runs phpunit, without codecoverage
phpunit --configuration tests/phpunit.xml
We use Travis-CI for continuous integration testing: each push to the Github repository triggers a build that executes PHP unit tests.
.travis.yml
tests/phunit.travis.xml
(and excludes code coverage reporting, to make the test suite run fast)
We're currently testing on PHP versions:
In order to skip a build by Travis, add the following to the commit message:
[ci skip]
With travis-lint, it's possible to check the configuration before deploying it: https://github.com/travis-ci/travis-lint
gem install travis-lint
travis-lint
We use codesniffer to test the guidelines according to a (old) zend guideline.
phpcs.sh runs the tests, we have excluded applicatation/models/base, because it is generated by doctrine
we use pdepend to test the complexity of the code
it results in:
Apart from codesniffer and pdepend, we use messdetector to get information about coding guidelines/complexity
phpmd.sh runs the tests
We are (going to) use duplicate code checker to check for duplicate code. It is an indicator that code is not reused proberly.
@todo script which runs the tests.
We run a test server at http://test4.webenq.org/tests featuring PHP 5.4 and Java 6.
Each night, it gets the latest version from Github and runs the complete testsuite
application/models
is currently excluded from code coverage reporting to make it fit into the 400MB memory of the (virtual) server (Amazon t1.micro).