rspec users - Jun 2011 - testing framework in 44 lines of ruby

Hi everyone,

I just came across this post and was wondering if this is on target with how
RSpec works under the hood?

http://www.skorks.com/2011/02/a-unit-testing-framework-in-44-lines-of-ruby/

Patrick J. Collins
http://collinatorstudios.com

On Jun 12, 2011, at 11:26 AM, Patrick J. Collins wrote:
> Hi everyone,
> 
> I just came across this post and was wondering if this is on target with
how
> RSpec works under the hood?
> 
> http://www.skorks.com/2011/02/a-unit-testing-framework-in-44-lines-of-ruby/
What do you mean by "on target"? Are you asking if the implementations
are the same, or similar? Or are you asking if attest meets the same goals as
RSpec?

If you''re asking about implementation, you can see both implementations
on github and judge for yourself:

https://github.com/rspec/rspec-core
https://github.com/rspec/rspec-expectations
https://github.com/rspec/rspec-mocks

https://github.com/skorks/attest

If you look at the latter, you''ll see that the blog post''s
title is at least a bit misleading. The
https://github.com/skorks/attest/blob/master/lib/attest directory has 5
subdirectories and 6 files, one of which
(https://github.com/skorks/attest/blob/master/lib/attest/execution_context.rb)
is close to 200 lines long. I''ll leave it to somebody else to do a more
thorough LOC comparison, but you get my drift.

Cheers,
David

> What do you mean by "on target"? Are you asking if the
implementations are
> the same, or similar? Or are you asking if attest meets the same goals as
> RSpec?
More specifically, I meant the way he implements describe blocks and the should
method...

It''s difficult for me to tell from glancing through the actual RSpec
code..
For example I am confused by the implementation of ''should'',
since it calls:
__should_for_example_group__

But prior to that, it appears that __should_for_example_group__ is an alias for
should, which looks to me like it would be recursively calling itself.

In that post''s example, their definition of should simply returns self
as they
described "it''s just syntactic sugar to make the tests read
better", but from
looking at RSpec''s code, it seems like there''s a lot more
going on, but I am
not quite seeing the big picture...

Patrick J. Collins
http://collinatorstudios.com

On Jun 12, 2011, at 12:52 PM, Patrick J. Collins wrote:
>> What do you mean by "on target"? Are you asking if the
implementations are
>> the same, or similar? Or are you asking if attest meets the same goals
as
>> RSpec?
> 
> More specifically, I meant the way he implements describe blocks and the
should
> method...
> 
> It''s difficult for me to tell from glancing through the actual
RSpec code..
> For example I am confused by the implementation of
''should'', since it calls:
> __should_for_example_group__
> 
> But prior to that, it appears that __should_for_example_group__ is an alias
for
> should, which looks to me like it would be recursively calling itself.
The code you''re looking at is an extension to support one liners like
this:

it { should do_something }

The original definition of should is in
https://github.com/rspec/rspec-expectations/blob/master/lib/rspec/expectations/extensions/kernel.rb,
which is what adds the ''should'' method to every object.
> In that post''s example, their definition of should simply returns
self as they
> described "it''s just syntactic sugar to make the tests read
better", but from
> looking at RSpec''s code, it seems like there''s a lot more
going on, but I am
> not quite seeing the big picture...
RSpec''s ''should'' actually does something: it accepts
a matcher and then tells the matcher to "match" against self. So no,
they are not doing the same thing.

I''m not really sure what you''re trying to evaluate here, but
feel free to ask if you have any more specific questions.

Cheers,
David

On Jun 12, 2011, at 10:52 AM, Patrick J. Collins wrote:
>> What do you mean by "on target"? Are you asking if the
implementations are
>> the same, or similar? Or are you asking if attest meets the same goals
as
>> RSpec?
> 
> More specifically, I meant the way he implements describe blocks and the
should
> method...
> 
> It''s difficult for me to tell from glancing through the actual
RSpec code..
Don''t glance at it! Study it :)

Pat