Re: Rethinking... (Re: RFC: fp printing speedup...)

[Martin Sebor <sebor at roguewave dot com>]

Nathan Myers wrote:

...
> > Well, wait.  What if ctype.widen looked like:
> >
> > CharT widen(char) {
> >  if (!cache) {
> > 	cache = getcache();
> > 	do_widen(cache.CharTarray, cache.chararry);
> >  }
> >  return cache.CharTarray[char];
> > }
> >
> > Would this help the speed?
>
>
>
> I don't think a ctype cache would help much, if at all.
> Better to avoid calling ctype<>::widen(char) too frequently.

That would be the ideal goal :) Our widen() looks similar to the
one above and it does make a significant difference in performance,
especially where virtual function calls are expensive.

>
...
> > If the standard does intend this to be the case, then we don't need to
> > loop with widen(char) and can use the array form instead.
>
>
> Absolutely you can use the array form.  Anybody who suggested that
> anything should work right when widen gives different results for 
> single char and array forms would be laughed off the list.

I'm sure Nathan is right but I agree that it would be nice if the
standard actually said so. I'll submit an issue. That said, it may
not be straightforward to eliminate the virtual call from the array
form. Plus, using the array form may also require another buffer to
copy to before inserting into the stream (although your implementation
has that snazzy ostreambuf_iterator to avoid this extra copy).

Martin