As declared, it won't actually save *any* space, because the union tag
will cause the size of the data section of your parent struct to grow to
one word (instead of zero, since everything else is pointers), so with
just the list pointer you have an extra word at the start of the struct
list vs. a direct pointer to the struct, whereas with the union you have
a bigger parent struct. If there were already other fields in the parent
struct such that adding the union didn't grow the struct, you might see
a marginal gain, but this feels like premature optimization to me.
See also the encoding spec:
https://capnproto.org/encoding.html
Quoting Cagatay K (2019-03-24 19:50:25)
> I have a case where a struct has a list of "child" structs, but the
> list almost always has a single element. Is there any benefit to using
> a union to switch between a list and a single element, like below?
> struct Parent {
> � name: Text;
> �
> � struct Child {
> � � id :UInt64;
> � � value: Text;
> � }
> � union {
> � � child :Child;
> � � children :List(Child);
> � }
> }
> The idea is that a single child would take less space, be faster to
> access, may be stored in a more compact way if it's small (for some
> definition of small), etc. Or is this a pointless exercise because even
> if there was an effect it would have been too small to matter?
> Cheers,
> �CK.
>
> --
> You received this message because you are subscribed to the Google
> Groups "Cap'n Proto" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to [1][email protected].
> Visit this group at [2]https://groups.google.com/group/capnproto.
>
> Verweise
>
> 1. mailto:[email protected]
> 2. https://groups.google.com/group/capnproto
--
You received this message because you are subscribed to the Google Groups
"Cap'n Proto" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
Visit this group at https://groups.google.com/group/capnproto.
