(j3.2006) Did we intend to prevent this?

[Bill Long]

On 11/10/11 3:47 AM, Robert Corbett wrote:
> The problem seems worse than Malcolm makes it out to be.  Assume a named
> constant has a pointer component that is initialized to be pointer associated
> with a variable.  Item (1) of the list given in Clause 7.1.12 (page 151) is
>
>        (1)  a constant or a subobject of a constant,
>
> As Val noted, a reference to a pointer component of a named constant will get
> the value of the variable that is the target of the pointer component.

I don't think this situation can occur if the original problem is fixed. 
  Constructors of a type that has component initialization, and in which 
the values of components that are default initialized to a non-constant 
object are not overridden with actual constants in the constructor 
simply should not qualify as constant expressions.

The current wording of 7.1.12 (3)(b) reads

  " a pointer component is an initialization target or a reference to 
the intrinsic function   NULL,"

The problem goes away of this is changed to

"a pointer component is a reference to the intrinsic function NULL,"

It was not clear what "initialization target" meant here anyway.  It 
would seem that the only thing that could qualify would be a reference 
to NULL() or a reference to a function that returned NULL().

Cheers,
Bill

> According to item (1), that reference to a pointer component can be a primary in
> a constant expression.  That is not good.
>
> Clause 16.5.2.3 does not cover the case of a target corresponding to an item in
> a structure constructor.
>
> Bob Corbett
> _______________________________________________
> J3 mailing list
> J3 at j3-fortran.org
> http://j3-fortran.org/mailman/listinfo/j3

-- 
Bill Long                                           longb at cray.com
Fortran Technical Support    &                 voice: 651-605-9024
Bioinformatics Software Development            fax:   651-605-9142
Cray Inc./Cray Plaza, Suite 210/380 Jackson St./St. Paul, MN 55101