Joe Buck wrote:
> On Wed, Apr 19, 2006 at 01:56:46PM -0700, Mark Mitchell wrote:
>> Let's accept that both the bit-preserving and value-preserving
>> conversions would be useful. How do we differentiate the two?
>>
>> In C++, we could invent __value_cast<T> and __bitwise_cast<T>. For
>> example, __bitwise_cast<int>(3.0f) would be defined as equivalent to:
>>
>> ({
>> union {
>> int i;
>> float f;
>> } temp;
>> temp.f = 3.0f;
>> temp.i;
>> })
>
> Those concepts roughly correspond to the existing static_cast and
> reinterpret_cast.
Indeed. (Although reinterpret_cast<int>(3.0f) is not valid C++, IIRC.)
However, I'd be very hesitant to break the correspondence between
static_cast and C-style casts, and it looks like the existing vector
specifications require that C-style casts behave like __bitwise_cast. A
C-style cast is defined as one of a const_cast, static_cast, or
reinterpret_cast, depending on the situation. So, it would be weird if
static_cast did something different than a C-style cast; hence the need
for __value_cast.
It might be that we could safely extend reinterpret_cast to handle this
case, though, making the only language addition __value_cast, which
would be there to work around the language impurity introduced by the
original vector specifications.
--
Mark Mitchell
CodeSourcery
[EMAIL PROTECTED]
(650) 331-3385 x713
