Re: List container strawman 1.3

["Mark Lundquist" <mlundquist2@attbi.com>]

"Ted Dennison" <dennison@telepath.com> wrote in message
news:%7sP7.49836$xS6.82296@www.newsranger.com...
> In article <3C0E1FA1.45A38A75@brighton.ac.uk>, John English says...
> >
> >Why does the iterator type need to be controlled? What's your mental
> >model of what an iterator consists of? Mine generally says that an
>
> An iterator doesn't but a *safe* iterator does (as the list needs to know
when
> its iterators go away).

Ted, why does a list need to "know" when its iterators go away?

Did you mean to say that safe iterator needs to know when its list has gone
away?

If so, a more lightweight approach is to use an access discriminant denoting
the list, e.g.:

    type Iterator (My_List : access List) is limited record
        My_Item : Item_Ptr;
    end record;

Then it is simply impossible for an Iterator to outlive its list!  The lack
of assignment is easily overcome:

    procedure Set (This : in out Iterator; To : Iterator) is
    begin
        This.My_Item := To.My_Item;
    end Set;

since composing ":=" is probably not an issue for iterators (i.e. the
ability to assign composite types having Iterator components).

Best,
Mark