Prev: awesome-cli-ruby
Next: Build 32 bit version of ruby 1.92 on snow leopard
From: Intransition on 18 Jun 2010 09:45
What kind of testing is it when you test a projects executables:
system/functional or integration?

From: Caleb Clausen on 18 Jun 2010 13:17
On 6/18/10, Intransition <transfire(a)gmail.com> wrote:
> What kind of testing is it when you test a projects executables:
> system/functional or integration?

These terms are so broad and mean so many things to so many people
that I think you can just take your pick.

From: R.. Kumar on 19 Jun 2010 01:05
Thomas Sawyer wrote:
> What kind of testing is it when you test a projects executables:
> system/functional or integration?

Is this about the terminology ?

I'd like to know what tools are being used for testing command-line
programs ?

I looked around to see samples of shoulda in command line gems but could
not find any. Currently, I am using a shell program (without a name or
repo) written by the git people (Junio Hamano) to test git. I used it to
test a shell app I wrote, and now I an using it for a ruby command line
app.

Sadly, it compares the standard output of a program - expected vs
actual. Which means every time I tweak or improve the output, my tests
fail. It ignores stderr. But otherwise it serves my purpose a lot -- far
better than no auto testing.

If this is OT, I could start a new thread for that.
--
Posted via http://www.ruby-forum.com/.

From: Intransition on 19 Jun 2010 06:14


On Jun 19, 1:05 am, "R.. Kumar" <sentinel1...(a)gmail.com> wrote:
> Thomas Sawyer wrote:
> > What kind of testing is it when you test a projects executables:
> > system/functional or integration?
>
> Is this about the terminology ?
>
> I'd like to know what tools are being used for testing command-line
> programs ?
>
> I looked around to see samples of shoulda in command line gems but could
> not find any. Currently, I am using a shell program (without a name or
> repo) written by the git people (Junio Hamano) to test git. I used it to
> test a shell app I wrote, and now I an using it for a ruby command line
> app.
>
> Sadly, it compares the standard output of a program - expected vs
> actual. Which means every time I tweak or improve the output, my tests
> fail. It ignores stderr. But otherwise it serves my purpose a lot -- far
> better than no auto testing.
>
> If this is OT, I could start a new thread for that.

No worries. I use either Cucumber (with Aruba) or QED.

And yes, checking the $stdout is part of test assertions, but I keep
it to specific snippets not the whole output, via
'out.index(pattern)'.

From: R.. Kumar on 19 Jun 2010 06:29
Thomas Sawyer wrote:

>>
>> If this is OT, I could start a new thread for that.
>
> No worries. I use either Cucumber (with Aruba) or QED.
>
> And yes, checking the $stdout is part of test assertions, but I keep
> it to specific snippets not the whole output, via
> 'out.index(pattern)'.

Is there some project you've used Cucumber or QED in (command-line). I
could see the test cases, and copy from there.
I didn't find much on the main page of QED. I did glance at Aruba some
days back but was lost.
--
Posted via http://www.ruby-forum.com/.

 |  Next  |  Last
Pages: 1 2
Prev: awesome-cli-ruby
Next: Build 32 bit version of ruby 1.92 on snow leopard