Re: Pointer usage danger?

["M. Lazarus" <mlazarus@xxxxxxxx>]

Barbara,

 The easy answer to several of the questions is that (unfortunately) 
this needs to compile back to v4r5. You're correct, of course, about 
the DIM() on the DS, but in the sample I was working with I actually 
had it defined as the first field and the other fields were using 
Overlay() and Overlay(*Next).  Many the techniques that I'd love to 
be able to use, just aren't available for this client.

 The idea behind this technique is that a screen is displayed to the 
user.  The user may want to reposition the view to any "record" any 
number of times.  An easy way to code this "window / view" would be 
to just reassign the pointer to the proper element when the user 
requests the reposition and (almost) no other code has to 
change.  Otherwise, the offset always needs to be taken into account, 
making the code harder to read, IMHO.

 I load Array1 with the full data set.  Array2 is then used to 
display and manipulate the data.  Therefore, the array index always = 
the subfile RRN.

 -mark


At 3/8/2011 01:36 PM, you wrote:
> On 2011/3/8 12:13 PM, M. Lazarus wrote:
> > Given this code:
> > ...
> > At this point, the first element of Array2 should be positioned over the
> > last element (#10) of Array1. Assuming I don't attempt to access element
> > 2 in Array2, is this safe to do or will the shift of Array2, elements
> > 2-10, overlay memory it shouldn't?
> >
>
> Mark, what's the purpose of having two arrays exactly the same except
> for the subfield names?
>
> By the way, to code the DIM keyword for a DS, you'll be forced to code
> the QUALIFIED keyword, so you won't have to worry about having the same
> subfield names in the two data structures.
>
> If you do need two data structure arrays for some reason, you'd be
> better off defining one of them with LIKEDS to guarantee you get the
> same layout.
>
> But I agree with others in this thread that you're setting up a
> dangerous situation by defining Array2 with the same number of elements
> as Array1, while intending to position the Array2 ypointer somewhere
> after the first element of Array1.  If you're only intending to use the
> first element of Array2 anyway, why make it an array?
>
> And now I wonder, why bother with the overlay in the first place?  Why
> not just use Array1(10).Field1 etc?