[boost] Re: Re: rational interface

From: Jonathan Turkanis (technews_at_[hidden])
Date: 2005-03-24 14:07:04

Andras Erdei wrote:
> what would you say to this:
>
>
> namespace rational
> {
>
>     template < class unlimited >
>   class unlimited
>     {
>       //...
>     } ;
>
>   enum policy
>     {
>       round ,
>       round_and_exactness ,
>       exception ,
>       assert
>     } ;
>
>   // for power users
>     template < typename limited , policy rounding = round >
>   class fxs_
>     {
>       //...
>     } ;
>
>   typedef __int64 biggest_builtin ;
>
>   // for end-users
>     template < biggest_builtin limit = INT_MAX , policy rounding =
>   round > class fxs
>       : public fxs_< fastest type that can hold [-limit...limit] ,
>     rounding > {
>       //...
>     } ;
>
>   // if you are really desperate and want to specify by hand
>   // the num_type, den_type, num_min, num_max, den_min, den_max etc
>   // there is a way to create your own policy blob
>
>   // of course we need better names than fxs and fxs_
>
> }

Do I understand correctly that rational::unlimited would be like the current
rational, while rational::fxs would provide a separate interface for rationals
based on built-in integral types?

Would there be any shared implementation?

Is fxs for power-users too, or just fxs_?

I agree that a better name than "fxs" is needed. Could you give me a hint what
it means? ;-)

Jonathan