Re: [boost] [optional] Reference assignment

From: Brian McNamara (lorgon_at_[hidden])
Date: 2003-10-31 11:52:48

On Fri, Oct 31, 2003 at 01:26:28PM +0800, Joel de Guzman wrote:
> Oh yes, you can always coerce an API to have it your way <wink ;>

So, if I understand your recent message, aside from nits, this does
work.  So...

> Give me a practical example that shows why your preferred semantics 
> is better than the tried and true semantics of true C++ references.
> I've given mine, following the premise that the match<T> is conceptually 
> an optional<T>. IMO, the match<T> perfectly fits the description of an 
> optional<T>. Hence, a use case.

Here is my use-case:

Inside Spirit, the match<T> class may or may not contain a value of
type T depending on whether or not the parse fails.  Thus we use an
optional<T> in its implementation.  Thanks to the clever current design
of optional, this enables match<T> to be implemented with this nice
generic code:

> >    template <class T>
> >    class match {
> >       optional<T> data;
> >       ...
> >    public:
> >       match( T x ) : data(x) {}
> >       ...
> >       void value( T y ) {
> >          *data = y;
> >       }
> >    };

Hurray for the current semantics of optional!

I know that sounds sarcastic, but seriously: is there anything wrong
with this use-case?  As far as I can tell, optional already works fine
for your needs.  (If it ain't broke, ...)

-- 
-Brian McNamara (lorgon_at_[hidden])