RE: [open-cobol-list] Re: NULL and EVALUATE

[Bill Klein]

Yes, not and medium <G> ....

I am in the process of having some discussions with J4 on this issue.  You
are correct (and I was wrong) that the '02 definitions of NULL (and they do
NOT allow NULLS - as IBM and MF do) is an identifier.  Therefore, it might
be valid as a selction subject or object.

HOWEVER, SR(1) on page 98 states,

"1) This format may be used only as a sending operand in an INITIALIZE or a
SET statement; as an argument in a program-prototype format CALL statement,
a function-prototype format function activation, or a method invocation; or
in a data-pointer or program-pointer relation-condition."

So the questin is whether use in a selection-object / selection-subject is
or is not a "relation-condition".

GR(4a6) on page 144 says

"6. If the selection object is identifier-2, literal-2,
arithmetic-expression-2, or boolean-expression-2, the pair is considered to
be a conditional expression of the following form:

selection-subject [NOT] = selection-object

where 'NOT' is present if it is present in the selection object. The result
of the analysis is the truth value of the resulting conditional expression."

but NORMALLY a GR doesn't override (clarify) a Syntax Rule.

  ***

Bottom-Line: (on this issue) 
   Is that as long as you are doing an '85 Standard compiler with an
EXTENSION, then you probably could/should allow
   EVALUATE NULL  *> or NULLS)
      When  Pointer-data-item

but there is still a question for the '02 Standard.

***

Also of interest (and I have clarified this) is the fact that the '02
Standard 

A) does NOT allow NULL (nulls) in a VALUE clause
B) *always* initializes a pointer data item to "NULL"  

> -----Original Message-----
> From: address@hidden 
> [mailto:address@hidden On 
> Behalf Of Roger While
> Sent: Monday, December 06, 2004 3:07 AM
> To: address@hidden
> Cc: address@hidden
> Subject: [open-cobol-list] Re: NULL and EVALUATE
> 
> Bill, I don't think I agree here.
> The definitions at 96/98 are under section 8.4.2 - Identifiers.
> This satisfies one of the conditions for the subject of an EVALUATE.
> The SR "in a data-pointer or program-pointer relation condition"
> says we can compare it, so I would interpret this as meaning it
> is perfectly valid as a subject of the EVALUATE.
> 
> Roger
> 
> 
> 
> >>First, I was in error (in an earlier post) when I indicated 
> that "NULL" 
> >>was
> >>a figurative constant.  This may (or may not) be true for 
> some extension 
> >>to
> >>the '85 Standard - but it is NOT true for the '02 Standard.
> >>
> >>Rather it is defined as "8.4.2.10 Predefined-address" (see page 98)
> >>    and/or
> >>  "predefined object reference" (see page 96)
> >>
> >>NEITHER of these is valid as the "Selection-subject" in an EVALUATE
> >>statement.
> >>
> >>See for example SR(1) on page 98 (for the NULL address)
> >>
> >>"1) This format may be used only as a sending operand in an 
> INITIALIZE or 
> >>a
> >>SET statement; as an argument in a program-prototype format CALL 
> >>statement,
> >>a function-prototype format function activation, or a 
> method invocation; 
> >>or
> >>in a data-pointer or program-pointer relation-condition."
> >>
> >>   ***
> 
> 
> 
> 
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from 
> real users.
> Discover which products truly live up to the hype. Start reading now. 
> http://productguide.itmanagersjournal.com/
> _______________________________________________
> open-cobol-list mailing list
> address@hidden
> https://lists.sourceforge.net/lists/listinfo/open-cobol-list
>