From: Jaakko Jarvi (jajarvi_at_[hidden])
Date: 2002-02-19 10:43:14
> >From the parsing standpoint, this is bad. It requires arbitrary lookahead to
> determine that a and b in the typeof expression are bound to the declarations
> of a and b that follow, and most parsing styles don't deal well with
> arbitrary lookahead.
And of course, the parameter names a and b should hide any global a and b.
Yes, the parsing is a problem.
template<class A, class B, parameter a, parameter b>
vector<typeof(a + b)>
sum(const vector<A>& a, const vector<B>& b)
No disease a new keyword wouldn't fix.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk