[computer-go] Don't cares in Pattern Matching

[Chris Fant]

On 6/14/06, Eric Boesch <ericboesch at hotmail.com> wrote:
> Pattern matching can require a time/space tradeoff. If the patterns are very
> irregular and full of holes, then an attempt to minimize time usage by
> treating each don't-care case that is tested as three or four different
> patterns could lead inevitably to space usage that grows expontentially with
> the number of patterns, as Chrilly alluded to.

Actually it grows linearly with the number of patterns and
exponentially with the average number of don't-cares per pattern.