- METHODS / EVENTS
Test::BDD::Cucumber::Harness - Base class for creating harnesses
Harnesses allow your feature files to be executed while telling the outside world about how the testing is going, and what's being tested. This is a base class for creating new harnesses. You can see Test::BDD::Cucumber::Harness::TermColor and Test::BDD::Cucumber::Harness::TestBuilder for examples, although if you need to interact with the results in a more exciting way, you'd be best off interacting with Test::BDD::Cucumber::Harness::Data.
Called at the start and end of feature execution respectively. Both methods accept a single argument of a Test::BDD::Cucumber::Model::Feature.
If you have a background section, then we execute it as a quasi-scenario step before each scenario. These hooks are fired before and after that, and passed in the Test::BDD::Cucmber::Model::Scenario that represents the Background section, and a a dataset hash (although why would you use that?)
Called at the start and end of scenario execution respectively. Both methods accept a Test::BDD::Cucmber::Model::Scenario module and a dataset hash.
Called at the start and end of step execution respectively. Both methods accept a Test::BDD::Cucmber::StepContext object.
step_done also accepts a Test::BDD::Cucumber::Model::Result object and an arrayref of arrayrefs with locations of consolidated matches, for highlighting.
[ [2,5], [7,9] ]
step_done, but for steps that have been called from other steps. None of the included harnesses respond to these methods, because generally the whole thing should be transparent, and the parent step handles passes, failures, etc.
Some tests will run one feature, some will run many. For this reason, you may have harnesses that have something they need to do on start (print an HTML header), that they shouldn't do at the start of every feature, or a close-down task (like running
done_testing()), that again shouldn't happen on every feature close-out, just the last.
$self as the single argument for both.
step_done with the step's result. Expected to silently add the result in to a pool that facilitate the
result method. No need to override this behaviour.
Returns a collective view on the passing status of all steps run so far, as a Test::BDD::Cucumber::Model::Result object. Default implementation should be fine for all your needs.
Copyright 2011-2016, Peter Sergeant; Licensed under the same terms as Perl