Re: Compile error - why?

[Hans Boldt <boldt@xxxxxxxxxx>]

Rich Duzenbury wrote:
> Uh, yeah, I kind of gathered.  I'm trying to figure out why (Hans?) such a
> restriction is in place.  It seems a bit silly.  Kind of like not being
> able to level break (cycle) on a binary field.
>
> I did find out that after V4R5 it gets easier, as the %char bif can
> supposedly take an additional parameter for the conversion format.
>
> At 05:30 AM 10/4/02, you wrote:
>
> > VARRYING is not allowed with this date conversion.
> > it should be fixed length character variable...
> >
> > Be ignorant than urban insane
> > Smile :-) from your Heart
> >
> > > -----Original Message-----
> > > From: Rich Duzenbury
> > [SMTP:rduz=6xRnKaLersmyXaTedYSxiUEOCMrvLtNR@public.gmane.org]
> > > Sent: Friday, October 04, 2002 2:51 AM
> > > To:   rpg400-l=Zwy7GipZuJhWk0Htik3J/w@public.gmane.org
> > > Subject:      Compile error - why?
> > >
> > > Hi,
> > >
> > > I'm getting a compile error on this.  What is that rationale for not
> > > allowing one to move from a date field into a VARYING alpha field?
> > This
> > > machine is on V4R5, btw.  Now it looks like I'll have to declare a
> > > non-varying field to accept the data, and then do an extra move or
> > eval to
> > > get the data over into $tagval.
> > >
> > > D work_date       S               D
> > > D $tagval         s            512    VARYING
> > > C     *usa          movel     work_date     $tagval
> > >
> > > *RNF7560 30      1 Variable-length field is not allowed with this
> > > operation.
> > >
> > > Curious.
Why is this restriction in place?

First, let's review the rules for using varying length character
variables in the fixed-form calc specs. When developing the varying
functionality, we agonized a while over the issue of how to provide
some sort of meaningful use of varying length variables in the calc
specs. It was a problem due to the weird semantics of the MOVE and
MOVEL operations (among others). For example, when moving a short
Factor 2 operand to a longer Result Field operand using MOVE, the
left-most characters are left unmodified. We tried coming up with
rules that made sense using varying length fields, but attempts
resulted in rather ornery and complex rules.

And so, we adopted the general principle that any varying length
variable coded in fixed-format calcs would be treated as a fixed
length variable field defined with the current length. One simple
rule - fixed-form calcs means varying length variables are treated
as fixed length.

So on to date variables. You know there's a rule that when moving a
character variable to a date variable, the character variable must
be long enough to hold a date value (or vice versa). The problem
with a varying length variable is that at compile time we don't know
if the variable will have a current length at run-time long enough
to hold a date value. And so we disallow all varying length
variables in date, time, and timestamp MOVE operations.

I suppose you could argue that we could add a check at run-time, and
I suppose that's true. But since there were alternatives, we decided
not to do that extra work. Now, with V5R1, there are even more
expression alternatives to do D/T/Z manipulation.

(Oh BTW, if you want to do level breaks on integer input fields (or
any other of the unsupported types), just define a character field
in that field position, and code the level indicator on that field.)

Cheers!  Hans