[J3] Question on function result characteristic in context of an extensibe type with SAME_TYPE_AS intrinsic

[Bill Long]

The type of the A argument to same_type_as is type(t).  The result type for the function new_t() is also type(t).  Since type(t) qualifies as extensible, the error message from compiler “another” seems bogus. Besides, there is no polymorphism here, so A and B pretty clearly have the same type. I don’t see a standard violation.

Cheers,
Bill

 
> On Jul 3, 2019, at 9:22 AM, Vipul Parekh via J3 <j3 at mailman.j3-fortran.org> wrote:
> 
> Consider the following:
> 
> module m
>   type :: t
>      integer :: i = 0
>   end type
> contains
>   function new_t( i ) result(n)
>      integer, intent(in) :: i
>      type(t) :: n
>      n%i = i
>   end function
>   function f( u ) result(l)
>      type(t), intent(in) :: u
>      logical :: l
>      l = same_type_as( A=u, B=new_t(i=42) )
>  end function
> end module
> 
> One processor compiles the above code ok whereas another one issue an
> error with the same_type_as intrinsic in connection with argument B
> not being an extensible type.
> 
> A question therefore: does the above code conform to the standard?
> 
> Thanks,
> Vipul Parekh

Bill Long                                                                       longb at cray.com
Principal Engineer, Fortran Technical Support &   voice:  651-605-9024
Bioinformatics Software Development                      fax:  651-605-9143
Cray Inc./ 2131 Lindau Lane/  Suite 1000/  Bloomington, MN  55425