Re: Re: syntax suggestion

On Thursday 04 March 2004 19:22, David Abrahams wrote:
> You can fix that by treating the result of the whole expression as an
> AST and evaluating it afterwards, but as I said, I worry about the
> compile-time costs.

Major cases are already pointed out by Daniel:

Daniel Wallin wrote:
> There are at least two cases where it won't work:
> 
> struct A { void f(); }
> struct B : A {};
> 
> class_<B>("B") [
> def("f", &B::f) // signature is void(A::*)() but A isn't
> // wrapped.
> ]
>
> and the most trivial..
> 
> class_<A>("A") [
> def(constructor<>()) // Error, we need to explicitly mention
> ]                        // the wrapped type for every constructor
> 

I agree that AST for the latter case cannot be eliminated. Though, I believe 
the former case can be represented internally as a template parameterizied 
only by a list of classes used in member pointers. The example exposes 
&B::foo which is actually &A::foo, therefore, mpl::list<A> would suffice. 
class_<B> needs mpl::list<A>for conversion from B* to A*. It's trivial to 
call A::foo after this conversion is applied.
Not the best but much better then AST of all class entities.
-- 
Alexander Nasonov
Independent Developer & Consultant

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQBAR5mPK8n/CZPtRx4RAh/QAJ9zqBluLMQ1GQR/NSyY9VkfmF+TWgCeJ9al
REnu2MbkAVPYhN9IpvEA/Sc=
=Qo8X
-----END PGP SIGNATURE-----