rspec

Comments and changes to this ticket

• 

  David Chelimsky December 7th, 2007 @ 04:07 AM

  • → State changed from “new” to “open”
  • → Assigned user changed from “” to “Dan North”

  That's actually as designed. Each block gets defined only once.

  Dan - WDYT?

• 

  Kero December 7th, 2007 @ 02:29 PM

  I see, when you would specify twice, which one to use afterwards... Neither?

• 

  David Chelimsky December 7th, 2007 @ 02:33 PM

  It uses the first definition and ignores any subsequent definitions. It always looks for a step definition and then uses it if it finds it. If it doesn't find one it creates one with the block you've supplied or, if no block is supplied, with a block that raises a PendingError.

• 

  Dan North December 7th, 2007 @ 05:58 PM

  I think it should fail with a DuplicateStepError or something. I want to know if I am trying to express the same intent (the step name) in two different ways. It almost certainly means that a) I forgot I already had one or b) the step name contains a hidden assumption that differs between the two cases.

  In the latter case I probably want to break out the assumption into a new Given.

  The one thing I don't want is for the same step name to mean different things at different times during the run. That way madness and ambiguity lies.

• 

  Kero December 7th, 2007 @ 07:44 PM

  OK, I agree with the observations.

  My blocks set up something very similar, hence I used the same name.

  a DuplicateStepError would have told me why things were not happening as I expected.

  Now how would a test for that look like...

  (When "again" do end).should raise_error DuplicateStepError

  or something?

  Shall I submit a new ticket?

  (adapting this one seems not right.)

• 

  David Chelimsky December 7th, 2007 @ 11:43 PM

  • → State changed from “open” to “resolved”

  A new ticket is fine. I'll go ahead and close this one now.

  Thanks.

• 

  David Chelimsky December 7th, 2007 @ 11:43 PM

  • → State changed from “resolved” to “invalid”