This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
When using the "Shoulda" testing framework (http://thoughtbot.com/projects/shoulda), NetBeans will run your Shoulda tests along with your traditional tests, but doesn't report on the Shoulda tests unless they fail. To repro: Create a project and install Shoulda (I'm doing the following with Ruby 1.8.7p72 and Rails 2.1.0): $ rails nb_shoulda $ cd nb_shoulda $ script/generate controller main index $ cd vendor/plugins $ git clone git://github.com/thoughtbot/shoulda.git shoulda $ cd ../.. Create a shoulda test in test/functional/main_controller.rb that looks like this: context "when index is called" do setup {get :index} should_respond_with :success end Run the functional tests: $ rake test:functionals Note the test count in the output: 2 tests, 2 assertions, 0 failures, 0 errors Now load that up as a project in NetBeans, and run the tests with NB's "Run/Debug Rake Task..." feature. Note the test count in the output: 1 tests, 0 failures, 0 errors NB is not counting the Shoulda test that passed. But it is running it, and it knows if it fails. You can verify by changing the Shoulda test to expect, for example, HTTP 403 from the index action so that it will fail: context "when index is called" do setup {get :index} should_respond_with :forbidden end Run the tests from within NB again, and note the count in the output. NetBeans is now reporting on the test, because it failed: 2 tests, 1 failures, 0 errors
Thanks for the report. Seems to be caused by that Shoulda uses a bit different format for notifying listeners registered for the TestCase::STARTED and FINISHED events. I have a fix ready, will push to the repo tomorrow or on Monday.
Should be fixed now, changeset #1973dfdea1ca. There will be an automatic notification in this issue when the fix has been integrated. Please let use know if you find any other cases where Shoulda tests are not working, it is very much possible that there are such cases as we haven't done much (any) testing with the Shoulda framework. If the fix works well, I will try to have it included in the first patch for 6.5.
Outstanding! Thanks for the fast turnaround. I'll keep you posted if I find anything more.
Integrated into 'main-golden', will be available in build *200810210201* on http://bits.netbeans.org/dev/nightly/ (upload may still be in progress) Changeset: http://hg.netbeans.org/main/rev/1973dfdea1ca User: Erno Mononen <emononen@netbeans.org> Log: #150613: NetBeans doesn't report "Shoulda" tests in its output unless they fail
This issue is really fixed on nighly builds. Thanks. I just don't understand why this didn't ship with NB6.5 final. Why do you plan to release this fix only on NB7.0M1?
> I just don't understand why this didn't ship with NB6.5 final. > Why do you plan to release this fix only on NB7.0M1? Unfortunately it was too late to fix this in 6.5, we were already in high resistance when the issue was first reported (only show stopper bugs can be fixed when in high resistance). If you can confirm that the issue is fixed and mark it as VERIFIED, we can include this in the next patch for 6.5.
Verified in nightly builds.
Thanks for verifying! Adding as a candidate for the next patch.
fix backported into release65_fixes branch http://hg.netbeans.org/release65_fixes/rev/cf4c925effad