The examples given on the site have the binary opIndex returning a value of the same type as the second parameter, as in

 struct A
	{
	    int opIndex(int i);
	    int opIndex(int i, int value);
}

Is this so that it can be reused in expressions such as the following

A a1 = ...
A a2 = ...

a1[10] = a2[5] = 3;

I presume the author returns the new value, rather than the previous value?

Matthew

"Matthew Wilson" <matthew.hat@stlsoft.dot.org> wrote in message news:brs079$2n2g$2@digitaldaemon.com...
> The examples given on the site have the binary opIndex returning a value
of
> the same type as the second parameter, as in
>
>  struct A
> {
>     int opIndex(int i);
>     int opIndex(int i, int value);
> }
>
> Is this so that it can be reused in expressions such as the following
>
> A a1 = ...
> A a2 = ...
>
> a1[10] = a2[5] = 3;
>
> I presume the author returns the new value, rather than the previous
value?

The return type and value from opIndex is whatever the programmer specifies it to be.