[rspec-devel] new pending feature (pending release)

Bob Cotton bob.cotton at rallydev.com
Wed Jun 27 13:13:27 EDT 2007


"David Chelimsky" <dchelimsky at gmail.com> writes:

> On 6/27/07, aslak hellesoy <aslak.hellesoy at gmail.com> wrote:
>> On 6/27/07, Dan North <dan at tastapod.com> wrote:
>> >
>> >  I think that's brilliant! Especially that the pending fails when the block
>> > works, so you have to remove the pending block before you can check in.
>> >
>> >  Mind if I steal it for rbehave?
>
> Why steal it? Just use it! You're already requiring parts of rspec,
> just require the necessary parts for this too :)

This is something that's been kicking around in the back of my mind
since I saw the rbehave announcement. Wasn't really sure where to
bring it up, here or on the rbehave dev list. Seems there's a larger
audience here, so here goes....

I can see some potential with the rbehave syntax, but over the last
few months I've spent quite some time submitting patches to rspec to
make it usable for functional testing: namely context_setup (which was
renamed to before(:all)) and shared behaviors. 

Those combined with the rich formatter framework and it's going to be
a while before rbehave catches up these regards, at least for us.

Is there any way to combine the two projects so that describe/it and
story/scenario/given/when/then are just alternative DSLs in a larger
framework?

- Bob

>
> If that's too complex, let me know and I'll make it so you can do
> something like require 'spec/expectations/pending' or something like
> that.
>
> David
>
>> >
>>
>> Of course not :0)
>>
>> Aslak
>>
>> >
>> >
>> >  Bob Cotton wrote:
>> >  "David Chelimsky" <dchelimsky at gmail.com> writes:
>> >
>> >
>> >
>> >  I just added a new "pending" feature. It's in trunk (rev 2118+) and
>> > will be part of the 1.0.6 release (*probably* within the next week).
>> >
>> > Read all about it:
>> > http://blog.davidchelimsky.net/articles/2007/06/23/pending-insert-reason-here
>> >
>> >  A very nice feature.
>> >
>> > Here's a twist to consider:
>> >
>> > We use rspec for functional testing of a large-ish
>> > application. Sometimes we find bugs that may take some time to be
>> > resolved, a few days, a few weeks. Once the bug is identified, I don't
>> > want to see it failing the nightly builds, day after day.
>> >
>> > pending() might be useful here, as I can say:
>> >
>> >  pending("Bug 12345 being fixed")
>> >
>> > But then I would have to wrap this in some conditional logic that
>> > tests if the bug is actually fixed.
>> >
>> > But, what if pending took a block:
>> >
>> >  pending("Bug 12345 being fixed") do
>> >  actual.should == expected
>> >  end
>> >
>> > and pending would fail (not with a pending exception) once the code
>> > inside the block started passing?
>> >
>> > That way we could mark something as a know issue, and be notified when
>> > that thing was fixed and go removed the pending.
>> >
>> > Thoughts?
>> >
>> > - Bob
>> > _______________________________________________
>> > rspec-devel mailing list
>> > rspec-devel at rubyforge.org
>> > http://rubyforge.org/mailman/listinfo/rspec-devel
>> >
>> >
>> >
>> > _______________________________________________
>> > rspec-devel mailing list
>> > rspec-devel at rubyforge.org
>> > http://rubyforge.org/mailman/listinfo/rspec-devel
>> >
>> _______________________________________________
>> rspec-devel mailing list
>> rspec-devel at rubyforge.org
>> http://rubyforge.org/mailman/listinfo/rspec-devel
>>
> _______________________________________________
> rspec-devel mailing list
> rspec-devel at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-devel


More information about the rspec-devel mailing list