Subj : Re: Why not unions
To   : comp.programming
From : Rob Thorpe
Date : Thu Jul 28 2005 02:12 am

Alex Fraser wrote:
> "Rob Thorpe" <robert.thorpe@antenova.com> wrote in message
> news:1122490853.570267.241310@f14g2000cwb.googlegroups.com...
> [snip]
> > In C there are two main uses of unions:
> > 1. To save space by using a single location to store many types of
> > data.
> [snip]
> > union bar {
> >   int x;
> >   double y;
> > }
> > struct foo {
> >   enum thing what_it_is;
> >   union bar b;
> > }
> >
> > foo.what_it_is describes what is in the union b, it could have values
> > DOUBLE  or INT for example.
> >
> > It's only worth making this a union if we need the space.  If we don't
> > then we could use:
> >
> > struct foo {
> >   enum thing what_it_is;
> >   int x;
> >   double y;
> >   char *z;
> > }
> >
> > Which is simpler and easier to check.
>
> Why do you think that?

If all foos are initiated to a known state, say 0 in the x,y and z
fields, then you can perform some checking on correctness of
foo.what_it_is.  If what_it_is says it's a double, but y is zero you
know it's wrong.  It can be harder if everything is in one pot, y is
say 5.63322678e16 which may be correct, or maybe a mislabeled char *.
To avoid the risk of mislabeling it's best to access these things
indirectly anyway.

> FWIW, I would generally use an "unnamed" (not sure if that is the correct
> term) union type declared inside the struct:
>
> struct foo {
>     enum thing what_it_is;
>     union {
>         int x;
>         /* etc */
>     } b;
> };

That's also a good way to do it, but sometimes it's useful to have the
union around to use by itself.

.