rspec users - Jul 2010 - newbie: how to preserve NoMethodError under stubbing?

Hi,

I''ve been browsing the RSpec book and the RDoc, but I can''t
see how to
ensure the following:

Stub an instance with a method it doesn''t have and raise NoMethodError
(or something like it.)

I want to do this because I don''t want to use incorrect names for my
dependency methods...

Thanks,

Lille

On Fri, Jul 23, 2010 at 6:46 PM, Lille <lille.penguini at gmail.com>
wrote:
> Hi,
>
> I''ve been browsing the RSpec book and the RDoc, but I
can''t see how to
> ensure the following:
>
> Stub an instance with a method it doesn''t have and raise
NoMethodError
> (or something like it.)
RSpec doesn''t support anything like that. I''m not sure if any
of the
Ruby frameworks do, though I''ve been involved with conversations about
this sort of thing before.

I wouldn''t want behaviour like this myself unless it could be invoked
explicitly with a command line argument, but was otherwise off. In
other words, you could do something like:

  rspec spec --audit-stubbed-methods

And then that would generate some sort of report.

Feel free to submit a feature request to
http://github.com/rspec/rspec-mocks if you''re interested in pursuing
such.

Cheers,
David

>
> I want to do this because I don''t want to use incorrect names for
my
> dependency methods...
>
> Thanks,
>
> Lille
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users
>

Lille,

On Sat, 24 Jul 2010 01:26 -0500, "David Chelimsky"
<dchelimsky at gmail.com> wrote:
> On Fri, Jul 23, 2010 at 6:46 PM, Lille <lille.penguini at gmail.com>
wrote:
> > Hi,
> >
> > I''ve been browsing the RSpec book and the RDoc, but I
can''t see how to
> > ensure the following:
> >
> > Stub an instance with a method it doesn''t have and raise
NoMethodError
> > (or something like it.)
> 
> RSpec doesn''t support anything like that. I''m not sure if
any of the
> Ruby frameworks do, though I''ve been involved with conversations
about
> this sort of thing before.
> 
> I wouldn''t want behaviour like this myself unless it could be
invoked
> explicitly with a command line argument, but was otherwise off. In
> other words, you could do something like:
> 
>   rspec spec --audit-stubbed-methods
> 
> And then that would generate some sort of report.
> 
> Feel free to submit a feature request to
> http://github.com/rspec/rspec-mocks if you''re interested in
pursuing
> such.
> 
> Cheers,
> David
Most people ask for this kind of feature from their unit tests because
they''re not combining mocking unit tests with end-to-end acceptance
tests. Are you using both?
> 
> 
> >
> > I want to do this because I don''t want to use incorrect names
for my
> > dependency methods...
> >
> > Thanks,
> >
> > Lille
> > _______________________________________________
> > rspec-users mailing list
> > rspec-users at rubyforge.org
> > http://rubyforge.org/mailman/listinfo/rspec-users
> >
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users
>

El 24/07/2010, a las 08:26, David Chelimsky escribi?:
> On Fri, Jul 23, 2010 at 6:46 PM, Lille <lille.penguini at gmail.com>
wrote:
>> Hi,
>> 
>> I''ve been browsing the RSpec book and the RDoc, but I
can''t see how to
>> ensure the following:
>> 
>> Stub an instance with a method it doesn''t have and raise
NoMethodError
>> (or something like it.)
> 
> RSpec doesn''t support anything like that. I''m not sure if
any of the
> Ruby frameworks do, though I''ve been involved with conversations
about
> this sort of thing before.
I''m not sure if I understand the request, but with RR you can do this:
>> require ''rr''
=> true
>> extend RR::Adapters::RRMethods
=> main
>> foo = Object.new
=> #<Object:0x101664bd8>
>> stub(foo).bar { raise NoMethodError }
=>
#<RR::DoubleDefinitions::DoubleDefinition:0x10165ce60>
>> foo.bar
NoMethodError: NoMethodError

Although like I said, not sure if I understood Lille''s request.

Cheers,
Wincent

On Jul 24, 2010, at 4:35 AM, Wincent Colaiuta wrote:
> El 24/07/2010, a las 08:26, David Chelimsky escribi?:
> 
>> On Fri, Jul 23, 2010 at 6:46 PM, Lille <lille.penguini at
gmail.com> wrote:
>>> Hi,
>>> 
>>> I''ve been browsing the RSpec book and the RDoc, but I
can''t see how to
>>> ensure the following:
>>> 
>>> Stub an instance with a method it doesn''t have and raise
NoMethodError
>>> (or something like it.)
>> 
>> RSpec doesn''t support anything like that. I''m not
sure if any of the
>> Ruby frameworks do, though I''ve been involved with
conversations about
>> this sort of thing before.
> 
> I''m not sure if I understand the request, but with RR you can do
this:
> 
>>> require ''rr''
> => true
>>> extend RR::Adapters::RRMethods
> => main
>>> foo = Object.new
> => #<Object:0x101664bd8>
>>> stub(foo).bar { raise NoMethodError }
> => #<RR::DoubleDefinitions::DoubleDefinition:0x10165ce60>
>>> foo.bar
> NoMethodError: NoMethodError
> 
> Although like I said, not sure if I understood Lille''s request.
I read the request as this:

class Foo
  def bar; end
end

f = Foo.new
f.stub(:barr)
#=> Error: "The Foo class does not have a ''barr''
method. Perhaps you meant to stub ''bar''"

For me this would have very limited utility because 1/2 the time I''m
deliberately stubbing methods that don''t even exist yet on doubles
doubling for objects that don''t even exist yet, which is why I
wouldn''t want it to be something that happens implicitly by default.

Here''s another thread on the same matter from a little over a year ago:

http://groups.google.com/group/rspec/browse_thread/thread/cf0b3eae192b9d8c/f5b77b008de628c7?lnk=gst&q=mock+audit#f5b77b008de628c7

Cheers,
David
> 
> Cheers,
> Wincent

David,

Yes, your reading of the request is what I originally meant.

I appreciate the points made by others in this thread and the thread
David has referred to.

As I continue to learn RSpec I will undoubtedly avail myself of the
approaches recommended above or in the linked thread, but I think an
argument to stub() per the following could be useful:

f.stub(:barr, :throw_no_such_method_error=>true)
#=> Error: "The Foo class does not have a ''barr''
method. Perhaps you
meant to stub ''bar''"

Thanks,

Lille

On Jul 24, 7:15?am, David Chelimsky <dchelim... at gmail.com>
wrote:
> On Jul 24, 2010, at 4:35 AM, Wincent Colaiuta wrote:
>
>
>
>
>
> > El 24/07/2010, a las 08:26, David Chelimsky escribi?:
>
> >> On Fri, Jul 23, 2010 at 6:46 PM, Lille <lille.pengu... at
gmail.com> wrote:
> >>> Hi,
>
> >>> I''ve been browsing the RSpec book and the RDoc, but I
can''t see how to
> >>> ensure the following:
>
> >>> Stub an instance with a method it doesn''t have and
raise NoMethodError
> >>> (or something like it.)
>
> >> RSpec doesn''t support anything like that. I''m
not sure if any of the
> >> Ruby frameworks do, though I''ve been involved with
conversations about
> >> this sort of thing before.
>
> > I''m not sure if I understand the request, but with RR you can
do this:
>
> >>> require ''rr''
> > => true
> >>> extend RR::Adapters::RRMethods
> > => main
> >>> foo = Object.new
> > => #<Object:0x101664bd8>
> >>> stub(foo).bar { raise NoMethodError }
> > => #<RR::DoubleDefinitions::DoubleDefinition:0x10165ce60>
> >>> foo.bar
> > NoMethodError: NoMethodError
>
> > Although like I said, not sure if I understood Lille''s
request.
>
> I read the request as this:
>
> class Foo
> ? def bar; end
> end
>
> f = Foo.new
> f.stub(:barr)
> #=> Error: "The Foo class does not have a ''barr''
method. Perhaps you meant to stub ''bar''"
>
> For me this would have very limited utility because 1/2 the time
I''m deliberately stubbing methods that don''t even exist yet on
doubles doubling for objects that don''t even exist yet, which is why I
wouldn''t want it to be something that happens implicitly by default.
>
> Here''s another thread on the same matter from a little over a year
ago:
>
> http://groups.google.com/group/rspec/browse_thread/thread/cf0b3eae192...
>
> Cheers,
> David
>
>
>
> > Cheers,
> > Wincent
>
> _______________________________________________
> rspec-users mailing list
> rspec-us... at
rubyforge.orghttp://rubyforge.org/mailman/listinfo/rspec-users

El 24/07/2010, a las 15:34, Lille escribi?:
> David,
> 
> Yes, your reading of the request is what I originally meant.
> 
> I appreciate the points made by others in this thread and the thread
> David has referred to.
> 
> As I continue to learn RSpec I will undoubtedly avail myself of the
> approaches recommended above or in the linked thread, but I think an
> argument to stub() per the following could be useful:
> 
> f.stub(:barr, :throw_no_such_method_error=>true)
> #=> Error: "The Foo class does not have a ''barr''
method. Perhaps you
> meant to stub ''bar''"
How about this:

  f.stub.existing(:bar)

That''s probably RR influencing me there, which employs things like
"stub.proxy" and so on. So maybe not such a good idea for rspec-mocks.

Cheers,
Wincent

On 24 Jul 2010, at 2:52 PM, Wincent Colaiuta wrote:
> How about this:
> 
>  f.stub.existing(:bar)
> 
> That''s probably RR influencing me there, which employs things like
"stub.proxy" and so on. So maybe not such a good idea for rspec-mocks.
Another option might be to deprecate #stub! as an alias for #stub and then bring
it back as an equivalent to `stub.existing` above.

Ash

-- 
http://www.patchspace.co.uk/
http://www.linkedin.com/in/ashleymoran