Re: test_strict_weak_ordering concept tester

From: rogeeff (rogeeff_at_[hidden])
Date: 2002-01-28 14:46:08

--- In boost_at_y..., "johnrpanzer" <jpanzer_at_n...> wrote:
> All,
> 
> Many moons ago, I asked if there would be any interest 
> in something to do runtime concept checking (at least 
> for ordering predicates).  One or two people expressed 
> some mild curiosity, so as part of my New Year's resolutions, 
> I've uploaded a draft implementation of an ordering 
> predicate checker to:
> 
> http://groups.yahoo.com/group/boost/files/test_strict_weak_ordering/
> 
> Example:
> 
> less_equal<T> is not a Strict Weak Ordering.  The code:
> 
>     int intdata[] = {3, 5, 8 , 29, 872, 5839};
>     assert(test_strict_weak_ordering(less_equal<int>(),
>                                      intdata,intdata+6));
> 
> will produce output like:
> 
>   Predicate fails irreflexivity; a<a not allowed:#0=3
>   Predicate fails irreflexivity; a<a not allowed:  #1=5
>   ...
>   Predicate fails antisymmetry; a<b and b<a:  #0=3, #0=3
>   Predicate fails antisymmetry; a<b and b<a:  #1=5, #1=5
>   ...
>   assertion failed, line 15: ...
> 
> 
> Please let me know what you think.
> 
> Regards,
> John Panzer
> jpanzer_at_n...
> jpanzer_at_a...

We could add this as unit test framework extention. But,
1. Why this algorithm?
2. What other algorithms we may consider?
3. Who are intended users for this?

Gennadiy.