Re: [gnugo-devel] dfa_unsort_1_25.1

[Gunnar Farneback]

Tanguy wrote:
> The most important to keep the incremental door open 
> is to keep a clear separation between the
> pure pattern matching and the constrain evaluation.

This is a good idea in any case. Except for the fact that there is
some complexity involved in storing callback data, I wouldn't mind if
the pattern matcher, at least optionally, could return an array of
callbacks, which the caller would be able to sort as it liked before
processing them.

Inge wrote:
> A long time ago, I proposed a keyword 'supercedes' for patterns.
> Consider the two patterns above (never mind pattern syntax here):

Then you must have called it something different, because I have
nothing like that in my mail archive.

> Pattern C1
> 
> XX
> ..
> XX
>  
> :supercedes(C2)
> 
> Pattern C2
> 
> X
> .
> X
> 
> This would mean that pattern C2 would never even be tried if C1
> matches.  I think that we could easily gain a lot of nodes here since
> we could easily create special case patterns for common, but expensive
> reading cases.

Anyway, this seems rather unmanageable to me. Also we already have a
cutoff in that the constraint of C2 won't be evaluated if the two
strings already have been amalgamated, e.g. by pattern C1.

/Gunnar