• Registered: 2010-03-02
  • Posts: 14

Topic: Weird testing cucumber behavior...

I have a scenario I've been working on that passes when I use the @wip tag. There are two other scenarios within the feature that  pass and a background. However, when I remove the @wip and run all my tests...the latest scenario fails.

Further weirdness. I add @wip tag to the top of the  feature, comment out the other two tests, it passes. Comment IN either one of the other two tests, it fails.

Further, further, weirdness, I use the @javascript tag and run all the tests and it PASSES, without @javascript, it fails. AND I AM NOT USING ANY JAVASCRIPT! (really). Nothing inserted dynamically, just rendering a partial.

In development, it does actually do what it says it is supposed to do...i.e. it passes.

Thoughts? I'm thinking it it might be a bug in culerity?

Edit: is there also a way in cucumber to freeze the db (debugging break) so I can take a look at its state at a certain point? It's difficult to figure it out from the test log.

Last edited by jordanp (2011-02-20 14:08:32)

  • Registered: 2010-03-02
  • Posts: 14

Re: Weird testing cucumber behavior...

I did  a diff on the test.log file running with and without the @javascript tag and the only main difference I could see happening was that when the javascript tag is used, it truncates the DB after each scenario, where as running normally does not...

Although it looks as though the 'normal' testing uses SAVEPOINTs but I'm not sure how those differ. The indexes definitely don't reset under normal circumstances while under @javascript they do.

Last edited by jordanp (2011-02-21 19:49:23)