Re: VARLEN dds field in RPG-IV

[rob@xxxxxxxxx]

Yes, the 2 byte area is implicitly defined.

If I have a variable length field defined to the program, and it's yet to 
be used, then
dsply %len(VarLen);  // should display 0
eval VarLen='Test';
dsply %len(VarLen);  // should display 4
/end-free
                          move  'Hi There'     VarLen
/free
dsply VarLen;  // should display 'Hi T'
/end-free
                          move  'X'            VarLen
/free
dsply VarLen;  // shold display 'Xi T'


Hey, if you want MOVE to do all of it's tricks with Move and MoveL and 
have it truncate, etc; then to allow this then for the compiler writers to 
do this they have to assume you know what you are doing with the move.


Rob Berendt
-- 
"They that can give up essential liberty to obtain a little temporary 
safety deserve neither liberty nor safety." 
Benjamin Franklin 





G Armour <garmour400r@xxxxxxxxx>
Sent by: rpg400-l-bounces@xxxxxxxxxxxx
08/29/2003 08:15 AM
Please respond to RPG programming on the AS400 / iSeries
 
        To:     RPG programming on the AS400 / iSeries 
<rpg400-l@xxxxxxxxxxxx>
        cc: 
        Fax to: 
        Subject:        Re: VARLEN dds field in RPG-IV


--- rob@xxxxxxxxx wrote:
> To check to see if the length was not set by the compiler would be as 
> difficult as checking to see if this was a divide by zero:
> /free
>  z=x/y;
> /end-free
> y could have been set in the program, could have been passed in as a 
> parameter, could have been read in from a file.  Same story with the 
> length.

But y would have been *defined*.  If I just reference a VARLEN field by 
name, is the
2-byte length area implicitly defined?  Probably the answer is yes, but 
that was the
debate I was having with myself.

> Would you be happy with a warning message
> RNF....  The MOVE opcode is about to be deprecated.  Find a better 
> alternative

LOL!  Yes, that would be great!

GA