rspec users - Apr 2012 - RSpec exactly behavior (it doesn't exactly fail)

Hiya all,

I''ve got a quick question related to RSpec. I was test-driving some
code and ended up in an endless loop. I was surprised by this, but traced it
down to the mock not failing on additional calls but only in the end. Let me
explain.

I was writing code like this:

   subject.wrapper.should_receive(:window_list).exactly(4).times.and_return {
       counter = counter + 1
       counter >= 4 ? [ "new window" ] : []
     }

The idea was that it would call the code-block 4 times exactly and then return a
new value (and thus stop calling it). As the code to implement wasn''t
there yet, it led to a recursive call. I had expected RSpec to stop after 4
calls though, as I had instructed the mock that I expected exactly 4 calls.

I added a new test in RSpec itself in precision_counts_spec.rb:

     it "fails when a method is called more than n times, but fails within
the method call" do
       @mock.should_receive(:random_call).exactly(1).times
       lambda do
         @mock.random_call
         @mock.random_call
       end.should raise_error(RSpec::Mocks::MockExpectationError)        
     end

which failed :( (or it failed to fail and therefore failed!)

It would be nice if it would fail. Is there any reason for not failing already
at this point in time?

(I''m using RSpec 2.6-0. I quickly browsed the latest and
didn''t see this changed)

Thanks,

Bas

On Sunday, April 15, 2012 at 1:32 AM, Bas Vodde wrote:
> 
> Hiya all,
> 
> I''ve got a quick question related to RSpec. I was test-driving
some code and ended up in an endless loop. I was surprised by this, but traced
it down to the mock not failing on additional calls but only in the end. Let me
explain.
> 
> I was writing code like this:
> 
> subject.wrapper.should_receive(:window_list).exactly(4).times.and_return {
> counter = counter + 1
> counter >= 4 ? [ "new window" ] : []
> }
> 
> The idea was that it would call the code-block 4 times exactly and then
return a new value (and thus stop calling it). As the code to implement
wasn''t there yet, it led to a recursive call. I had expected RSpec to
stop after 4 calls though, as I had instructed the mock that I expected exactly
4 calls.
> 
> I added a new test in RSpec itself in precision_counts_spec.rb:
> 
> it "fails when a method is called more than n times, but fails within
the method call" do
> @mock.should_receive(:random_call).exactly(1).times
> lambda do
> @mock.random_call
> @mock.random_call
> end.should raise_error(RSpec::Mocks::MockExpectationError) 
> end
> 
> which failed :( (or it failed to fail and therefore failed!)
> 
> It would be nice if it would fail. Is there any reason for not failing
already at this point in time?
> 
> (I''m using RSpec 2.6-0. I quickly browsed the latest and
didn''t see this changed)
> 
> Thanks,
> 
> Bas
There is no philosophical reason for this to happen, and there are other types
of failures that do fail-fast (e.g. obj.should_receive(:bar).with(1,2) fails
immediately if it receives :bar with any other args).

Please submit this to https://github.com/rspec/rspec-mocks/issues and
I''ll start looking into a fix.

Cheers,
David
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://rubyforge.org/pipermail/rspec-users/attachments/20120415/936ae615/attachment.html>

Actually I just went ahead and fixed it sans-bug report:
https://github.com/rspec/rspec-mocks/commit/fb9c76c2e40b4b25f4dcc5de95f8c60319b6d9c1.
It''ll be fixed in the next release (2.10).

Cheers,
David


-- 
David Chelimsky
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)


On Sunday, April 15, 2012 at 4:48 PM, David Chelimsky wrote:
> On Sunday, April 15, 2012 at 1:32 AM, Bas Vodde wrote:
> > 
> > Hiya all,
> > 
> > I''ve got a quick question related to RSpec. I was
test-driving some code and ended up in an endless loop. I was surprised by this,
but traced it down to the mock not failing on additional calls but only in the
end. Let me explain.
> > 
> > I was writing code like this:
> > 
> >
subject.wrapper.should_receive(:window_list).exactly(4).times.and_return {
> > counter = counter + 1
> > counter >= 4 ? [ "new window" ] : []
> > }
> > 
> > The idea was that it would call the code-block 4 times exactly and
then return a new value (and thus stop calling it). As the code to implement
wasn''t there yet, it led to a recursive call. I had expected RSpec to
stop after 4 calls though, as I had instructed the mock that I expected exactly
4 calls.
> > 
> > I added a new test in RSpec itself in precision_counts_spec.rb:
> > 
> > it "fails when a method is called more than n times, but fails
within the method call" do
> > @mock.should_receive(:random_call).exactly(1).times
> > lambda do
> > @mock.random_call
> > @mock.random_call
> > end.should raise_error(RSpec::Mocks::MockExpectationError) 
> > end
> > 
> > which failed :( (or it failed to fail and therefore failed!)
> > 
> > It would be nice if it would fail. Is there any reason for not failing
already at this point in time?
> > 
> > (I''m using RSpec 2.6-0. I quickly browsed the latest and
didn''t see this changed)
> > 
> > Thanks,
> > 
> > Bas
> There is no philosophical reason for this to happen, and there are other
types of failures that do fail-fast (e.g. obj.should_receive(:bar).with(1,2)
fails immediately if it receives :bar with any other args).
> 
> Please submit this to https://github.com/rspec/rspec-mocks/issues and
I''ll start looking into a fix.
> 
> Cheers,
> David
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://rubyforge.org/pipermail/rspec-users/attachments/20120415/71511e24/attachment.html>

Hiya,

Kewl, thanks!

(To be honest, I was a bit disapointed as I was thinking of doing it myself and
sending it to you :P)

Anyways, much appreciated!

Bas

On 16-Apr-2012, at 10:14 AM, David Chelimsky wrote:
> Actually I just went ahead and fixed it sans-bug report:
https://github.com/rspec/rspec-mocks/commit/fb9c76c2e40b4b25f4dcc5de95f8c60319b6d9c1.
It''ll be fixed in the next release (2.10).
> 
> Cheers,
> David
> 
> -- 
> David Chelimsky
> Sent with Sparrow
> 
> On Sunday, April 15, 2012 at 4:48 PM, David Chelimsky wrote:
> 
>> On Sunday, April 15, 2012 at 1:32 AM, Bas Vodde wrote:
>>> 
>>> Hiya all,
>>> 
>>> I''ve got a quick question related to RSpec. I was
test-driving some code and ended up in an endless loop. I was surprised by this,
but traced it down to the mock not failing on additional calls but only in the
end. Let me explain.
>>> 
>>> I was writing code like this:
>>> 
>>>
subject.wrapper.should_receive(:window_list).exactly(4).times.and_return {
>>> counter = counter + 1
>>> counter >= 4 ? [ "new window" ] : []
>>> }
>>> 
>>> The idea was that it would call the code-block 4 times exactly and
then return a new value (and thus stop calling it). As the code to implement
wasn''t there yet, it led to a recursive call. I had expected RSpec to
stop after 4 calls though, as I had instructed the mock that I expected exactly
4 calls.
>>> 
>>> I added a new test in RSpec itself in precision_counts_spec.rb:
>>> 
>>> it "fails when a method is called more than n times, but fails
within the method call" do
>>> @mock.should_receive(:random_call).exactly(1).times
>>> lambda do
>>> @mock.random_call
>>> @mock.random_call
>>> end.should raise_error(RSpec::Mocks::MockExpectationError)
>>> end
>>> 
>>> which failed :( (or it failed to fail and therefore failed!)
>>> 
>>> It would be nice if it would fail. Is there any reason for not
failing already at this point in time?
>>> 
>>> (I''m using RSpec 2.6-0. I quickly browsed the latest and
didn''t see this changed)
>>> 
>>> Thanks,
>>> 
>>> Bas
>> There is no philosophical reason for this to happen, and there are
other types of failures that do fail-fast (e.g.
obj.should_receive(:bar).with(1,2) fails immediately if it receives :bar with
any other args).
>> 
>> Please submit this to https://github.com/rspec/rspec-mocks/issues and
I''ll start looking into a fix.
>> 
>> Cheers,
>> David
> 
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users

Sorry about that. I don''t always jump on things like that, but this one
was bugging me and I had the time so I just did it. That doesn''t always
happen :) I look forward to your future contributions.

Cheers,
David


-- 
David Chelimsky
Sent with Sparrow (http://www.sparrowmailapp.com/?sig)


On Sunday, April 15, 2012 at 9:30 PM, Bas Vodde wrote:
> 
> Hiya,
> 
> Kewl, thanks!
> 
> (To be honest, I was a bit disapointed as I was thinking of doing it myself
and sending it to you :P)
> 
> Anyways, much appreciated!
> 
> Bas
> 
> On 16-Apr-2012, at 10:14 AM, David Chelimsky wrote:
> 
> > Actually I just went ahead and fixed it sans-bug report:
https://github.com/rspec/rspec-mocks/commit/fb9c76c2e40b4b25f4dcc5de95f8c60319b6d9c1.
It''ll be fixed in the next release (2.10).
> > 
> > Cheers,
> > David
> > 
> > -- 
> > David Chelimsky
> > Sent with Sparrow
> > 
> > On Sunday, April 15, 2012 at 4:48 PM, David Chelimsky wrote:
> > 
> > > On Sunday, April 15, 2012 at 1:32 AM, Bas Vodde wrote:
> > > > 
> > > > Hiya all,
> > > > 
> > > > I''ve got a quick question related to RSpec. I was
test-driving some code and ended up in an endless loop. I was surprised by this,
but traced it down to the mock not failing on additional calls but only in the
end. Let me explain.
> > > > 
> > > > I was writing code like this:
> > > > 
> > > >
subject.wrapper.should_receive(:window_list).exactly(4).times.and_return {
> > > > counter = counter + 1
> > > > counter >= 4 ? [ "new window" ] : []
> > > > }
> > > > 
> > > > The idea was that it would call the code-block 4 times
exactly and then return a new value (and thus stop calling it). As the code to
implement wasn''t there yet, it led to a recursive call. I had expected
RSpec to stop after 4 calls though, as I had instructed the mock that I expected
exactly 4 calls.
> > > > 
> > > > I added a new test in RSpec itself in
precision_counts_spec.rb:
> > > > 
> > > > it "fails when a method is called more than n times,
but fails within the method call" do
> > > > @mock.should_receive(:random_call).exactly(1).times
> > > > lambda do
> > > > @mock.random_call
> > > > @mock.random_call
> > > > end.should raise_error(RSpec::Mocks::MockExpectationError)
> > > > end
> > > > 
> > > > which failed :( (or it failed to fail and therefore failed!)
> > > > 
> > > > It would be nice if it would fail. Is there any reason for
not failing already at this point in time?
> > > > 
> > > > (I''m using RSpec 2.6-0. I quickly browsed the
latest and didn''t see this changed)
> > > > 
> > > > Thanks,
> > > > 
> > > > Bas
> > > There is no philosophical reason for this to happen, and there
are other types of failures that do fail-fast (e.g.
obj.should_receive(:bar).with(1,2) fails immediately if it receives :bar with
any other args).
> > > 
> > > Please submit this to https://github.com/rspec/rspec-mocks/issues
and I''ll start looking into a fix.
> > > 
> > > Cheers,
> > > David
> > > 
> > 
> > 
> > _______________________________________________
> > rspec-users mailing list
> > rspec-users at rubyforge.org (mailto:rspec-users at rubyforge.org)
> > http://rubyforge.org/mailman/listinfo/rspec-users
> > 
> 
> 
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org (mailto:rspec-users at rubyforge.org)
> http://rubyforge.org/mailman/listinfo/rspec-users
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://rubyforge.org/pipermail/rspec-users/attachments/20120415/3b696ffa/attachment.html>