[RFC] Two puzzling things in num_put::do_put(void*)

[Paolo Carlini]

Hi,

the most important one is the presence of a try/catch construct:

  __io.flags(__flags & __fmt | (ios_base::hex | ios_base::showbase));
  try
    {
      __s = _M_insert_int(__s, __io, __fill,
                          reinterpret_cast<unsigned long>(__v));
      __io.flags(__flags);
    }
  catch (...)
    {
      __io.flags(__flags);
      __throw_exception_again;
    }

which is not present in num_get::do_get(void*&), or in any other
num_get/num_put virtual for that matter. Indeed, I don't find
support for its presence in the standard...

Less important, I think there is the hidden assumption that
sizeof(void*) == sizeof(unsigned long). This is probably true (on
64 bit platforms too, if I understand well some notes prepared by
Andreas) but I'm not sure if it's something we can really be 100%
confident about.

Paolo.