From: Jim Argeropoulos (jaa_at_[hidden])
Date: 2000-01-16 15:49:46
> This is the fault of the definition of
> logical_and/logical_not. Actually, most of std functors have this
> problem, I agree it is annoying. Perhaps the thing to do is create
> versions of the std functors that allow the operands to be different
> types.
OK, I played with this and short of putting a "new" version in the boost
namespace, the compiler won't take it.
But here is my real reason for writing.
With the file I proposed I can make the following typedef for a set of
Person sorted by three attributes.
typedef set < Person* , GreaterMemberFunc < Person , long , string , string
> > PersonSet ;
If I proceede and make my to use a new definition of boost::logical_or it is
going to look something like
typedef set < Person* , compose_f_gx_hy_t < boost::logical_or <
compose_f_gx_hy_t < less < long > , const_mem_func_t <
Well I'll quit there. It is going to be a mess that I don't ever want to
type out. I tried thinking out some typedef magic that could help, but I
gave up.
While I like the cool reusableness of the concept. It seems like it fails in
the above example.
Any suggestions on typedef magic that could help?
Jim