test for lib.set.difference

[Anton Pevtsov <an...@moscow.vdiweb.com>]

The attached file contains the test for the lib.set.difference
algorithm. 

With best wishes,
Anton Pevtsov

Re: test for lib.set.difference

[Martin Sebor <se...@roguewave.com>]

Anton Pevtsov wrote:
> The attached file contains the test for the lib.set.difference
> algorithm.

Great! Committed thus:
http://svn.apache.org/viewcvs.cgi?rev=378582&view=rev

Martin

Re: 23.vector.cons.cpp

[Martin Sebor <se...@roguewave.com>]

Liviu Nicoara wrote:
> I have converted the former vector.cpp to 23.vector.cons.cpp. I kept in
> it the types and signatures tests along with the actual ctors tests.

Okay, thanks! Committed thus:
http://svn.apache.org/viewcvs?rev=379235&view=rev

It would be good to enhance the test to exercise the template
specialized on class X (instead of int or short) and the ctor
template on the remaining iterator categories in addition to
InputIterators.

Also, if we go with the new name and reduced scope for the test
we should move the tests for vector::iterator to a file of their
own.

Finally, ideally, I would like to rewrite this and the rest of
our tests along the same lines as most of the algorithms tests,
i.e., as a series of small isolated hardcoded test cases, and
avoid algorithmically generating them. The generated tests
tend to be harder to read and, IMO, do not exercise the tested
functionality as well as handcrafted ones can (especially not
the corner cases). Changing the behavior of these tests also
tends to be much more difficult than in the handcrafted case.

Martin

23.vector.cons.cpp

[Liviu Nicoara <ni...@roguewave.com>]

I have converted the former vector.cpp to 23.vector.cons.cpp. I kept in
it the types and signatures tests along with the actual ctors tests.

Liviu