Ah, makes sense. Thanks!
On Saturday, April 26, 2014 12:33:48 PM UTC-4, Tim Holy wrote:
>
> When x is of type T, convert(T,x) is a no-op:
>
> julia> function myfunc(x)
> a = convert(Float64, x)
> return a
> end
> myfunc (generic function with 1 method)
>
> julia> code_native(myfunc, (Float32,))
> .text
> Filename: none
> Source line: 2
> push RBP
> mov RBP, RSP
> Source line: 2
> movabs RAX, 140048916337192
> movss DWORD PTR [RAX], XMM0
> movss XMM0, DWORD PTR [RAX]
> cvtss2sd XMM0, XMM0
> Source line: 3
> pop RBP
> ret
>
> julia> code_native(myfunc, (Float64,))
> .text
> Filename: none
> Source line: 3
> push RBP
> mov RBP, RSP
> Source line: 3
> pop RBP
> ret
>
>
> --Tim
>
> On Saturday, April 26, 2014 08:45:21 AM Jarrett Revels wrote:
> > Hello all,
> >
> > I was exploring some of the base code today when I saw the
> implementation
> > of setindex!():
> >
> >
> > setindex!{T}(A::Array{T}, x, i0::Real) = arrayset(A, convert(T,x),
> > to_index(i0))
> >
> > I was surprised that I couldn't find setindex!() methods for elements
> that
> > used the type system to forego the use of convert() seen above. For
> > example, wouldn't it be better to also define methods of the form
> >
> > setindex!{T}(A::Array{T}, x::T, i0::Real) = arrayset(A, x, to_index(i0))
> >
> > so that conversion only occurs when necessary?
> >
> > I would be willing to make the changes myself, but I assume that there
> is a
> > reason for this design choice that I'm just seeing.
>
