Den måndag 19 oktober 2015 kl. 10:28:15 UTC+2 skrev Uliano Guerrini:
>
>
>
> Il giorno lunedì 19 ottobre 2015 08:56:32 UTC+2, Gunnar Farnebäck ha
> scritto:
>>
>> It might help if you explain something about what you want to do with the
>> binary representation.
>>
>
> I would like to have an efficient way to access bytes, 16 and 32 bit words
> of larger types (in this case 8 bytes, not necessarily floating point).
> Possibly as efficient and as easy as it could be in C/C++
>
> struct Words32 {
>
> int32_t lo;
>
> int32_t hi;
>
> };
>
> union Both {
>
> struct Words32 w;
>
> int64_t i;
>
> };
>
> union Both b;
>
> b.i=0x7fffffff3fffffff;
>
> printf("lo:0x%08x\nhi:0x%08x\n",b.w.lo,b.w.hi);
>
> gives
>
> *lo:0x3fffffff*
>
> *hi:0x7fffffff*
>
> *type Words32*
>
> *hi::Int32*
>
> *lo::Int32*
>
> *end*
>
> *a=0x7fffffff3fffffff*
>
> *reinterpret(Words32,a)*
>
>
> *gives*
>
>
> *ERROR: reinterpret: expected bits type as first argument*
>
> * in reinterpret at essentials.jl:115*
>
>
>
>
>>
>> For example,
>>
>> julia> f(x::Float64) = reinterpret(Float64, reinterpret(UInt64, x) &
>> 0x7fffffffffffffff)
>> f (generic function with 1 method)
>>
>
>
> I have seen that I can reinterpret a bits type into another bits type of
> the SAME size. unfortunately my Words32 isn't, even if it size is
> completely specified and should be known ahead of time by the compiler. I'm
> looking for ways to access its smaller chunks without passing through a
> conversion into an hexadecimal string.
>
Is this good enough?
julia> immutable Words32
hi::Int32
lo::Int32
end
julia> a = 0x7fffffff3fffffff
0x7fffffff3fffffff
julia> reinterpret(Words32, [a])
1-element Array{Words32,1}:
Words32(1073741823,2147483647)
