[boost] Re: char_separator and offset_separator speed

From: Robert Zeh (razeh_at_[hidden])
Date: 2004-01-05 17:27:39

Beman Dawes <bdawes_at_[hidden]> writes:

> At 11:43 AM 12/3/2003, Robert Zeh wrote:
>  >
>  >I have completed implementations of char_separator and
>  >offset_separator that are many times faster then the current
>  >implementations.  They are faster because they use the Token's assign
>  >method rather then operator +=.
>  >
>  >When supplied with input_iterators the faster implementations fall
>  >back to operator +=, and there is no speed gain.
>  >
>  >...
>  >
>  >A more description and some timing code is available at:
>  >
>  >http://home.earthlink.net/~rzeh/Fast_boost__tokenizer_tokenizer_function.html
> 
> Hi Robert,
> 
> Looks interesting.
> 
> I'm not sure that John Bandela reads this list regularly, so please
> try to contact him directly if you don't get a response here.
> 
> If that fails, post something here and we'll round up someone else to
> review you suggested changes.
> 
> Thanks,
> 
> --Beman

John R. Bandela has a proposed fix that he should be ready to check
into CVS, but it uses boost::mpl.  Does this new dependency bother anyone?

The fix uses boost::mpl so that it can be used with iterators that are
pointers on compilers that do not support partial template
specialization. 

Thanks,
Robert