> Why speculate, when you can instrument...ate?
> Conclusions:
> 1. J�rg's method of using nested greps works really well for
> small sets, but degrades poorly.
> 2. Hashes degrade much more gracefully.
> 3. Adam's method of manually iterating through both sets
> degrades linearly, suggesting that it is bound by the
> performance of the Perl interpreter, not the data structures.
> 4. Bitfields, at least as I have implemented them, fall
> somewhere between nested greps and hashes. I suspect the
> explicit loops required to populate, manipulate, and extract the
> sets from the bitfield.
Other advantage to hashes: non-numeric (don't see a way for
bitfields to handle strings gracefully).
sl
