(j3.2006) Is this prohibited? (with obvious blunder repaired)

Robert Corbett robert.corbett
Thu Jun 20 04:15:23 EDT 2013


On 06/19/13 19:37, Malcolm Cohen wrote:
>> I agree with Van.
> Ah, so you've implemented
>     REAL(KIND=MERGE(4,8,PRESENT(X))) Y
> then?

No,we have not.

I understand your point.  I agree with Van that the expression

MERGE(4,8,PRESENT(X))

is a specification expression according to the Fortran 2008 standard.  I agree 
with you that it is also a constant expression according to the standard.  I 
think the error is in the definition of a constant expression, not in the 
definition of a specification expression.

Robert Corbett




More information about the J3 mailing list