Opinions Wanted: List API loop logic

[Scott Klement <klemscot@xxxxxxxxxxxx>]

First, some background:

In this morning's "Club Tech iSeries Programming Tips" newsletter, I
listed the following code snippet:

     c                   eval      p_ListEntry = p_ListHeader + DataOffset

     c                   for       EntryNo = 1 to NumEntries

     c***       ... at this point, the JOBL0100 data structure contains
     c***       ... one of the entries of the list.

     c                   eval      p_ListEntry = p_ListEntry + EntrySize
     c                   endfor

This code is intended to run through a list of entries in a user space as
a demonstration of how to use the list APIs.

Now, because I have my pointer incremented at the end of the loop, it'll
always do it one extra time.  Barbara tells me that it's possible that
p_ListEntry could be set to an illegal location, causing an MCH0601 error.
(Before reading her note, I had thought that as long as I didn't use the
data that the pointer points to after the loop ends, I was safe.)

Even if the user space is set to auto-extend, it could still cause a
problem because there's a limit to how large a user space might be, and I
might happen to hit that limit exactly.  (okay, it's not all that likely,
but still possible...)

Now my question:

Which of the following code samples do you think is the best way to do it,
and why?

This is the way I used to write this sort of loop, but I've found that
maintenance programmers don't like the way the pointer logic works, it
confuses them, plus it's ugly when wrapped to two-lines:

      c                   eval      p_ListStart = p_ListHeader + DataOffset

      c                   for       EntryNo = 1 to NumEntries
      c                   eval      p_ListEntry = p_ListStart +
      c                                     ( EntrySize * (EntryNo - 1) )

      c***       ... at this point, the JOBL0100 data structure contains
      c***       ... one of the entries of the list.

      c                   endfor


Here's another alternative, where the pointer work is done at the start of
the loop... but since it already starts on the first entry of the list,
it's necessary to skip the pointer math the first time through.  I don't
know how intuitive it is to do it this way:

    c                   eval      p_ListEntry = p_ListHeader + DataOffset

    c                   for       EntryNo = 1 to NumEntries

    c                   if        EntryNo <> 1
    c                   eval      p_ListEntry = p_ListEntry + EntrySize
    c                   endif

    c***       ... at this point, the JOBL0100 data structure contains
    c***       ... one of the entries of the list.

    c                   endfor

Finally, here's a similar example to the last one, except that it sets the
first entry inside the loop as well as the others:


    c                   for       EntryNo = 1 to NumEntries

    c                   if        EntryNo = 1
    c                   eval      p_ListEntry = p_ListHeader + DataOffset
    c                   else
    c                   eval      p_ListEntry = p_ListEntry + EntrySize
    c                   endif

    c***       ... at this point, the JOBL0100 data structure contains
    c***       ... one of the entries of the list.

    c                   endfor

So, what does everyone think?  Which one is the most intuitive of the
bunch?  Or, does anyone have a better idea?

Thanks!