(j3.2006) variables in DATA statements

[Malcolm Cohen]

Since the standard does not establish an interpretation for these (the user 
function is never invoked, as it is neither in a specification expression nor an 
executable construct, so there is no variable specified, and furthermore the 
variable is initially defined i.e. before invoking anything were anything to be 
invoked), it seems to be already prohibited by our means of last resort.  That 
is to say, in an unsatisfactory manner.

I thought briefly that C568 might have done the job, but as currently worded it 
does not.  Perhaps it should be reworded something like
   "If a thingo that appears as a data-stmt-object is a variable, it shall be an 
object designator in which no pointer appears unless it is the entire rightmost 
part-ref."

Clearly an interp is called for to resolve the unsatisfactory way this is 
handled.

-----Original Message----- 
From: Robert Corbett
Date: ?? 24?4?12? 9:58
To: j3 at j3-fortran.org
Subject: (j3.2006) variables in DATA statements

The syntax rules for DATA statements in Fortran 2008
(see Clause 5.4.7, page 104) allow variables that
have the form of references to functions as
/data-stmt-object/s.  I find nothing in the
description of DATA statements to prohibit such
variables from appearing in DATA statements.  I
assume that such variables should not be permitted
in DATA statements.  Does anyone disagree?

Robert Corbett
_______________________________________________
J3 mailing list
J3 at j3-fortran.org
http://j3-fortran.org/mailman/listinfo/j3

________________________________________________________________________
This e-mail has been scanned for all viruses by Star.
________________________________________________________________________

-- 
................................Malcolm Cohen, Nihon NAG, Tokyo.