* Am 12.03.06 schrieb Conor McBride:
> funny-pattern-matching twists. One is that I built a flexOrRigid case
> analysis which analyses a term as either (var x) or t in such a way
> that subsequently doing case t doesn't give you a var case ('cos you
> already know it's rigid).
And it seems to do so without inspecting the n of t : Tm n , as i
naïvely thaught when first reading this. Like that a t from a language
with no variables Tm zero would give you just rigid cases.
And thank's for posting it. As you might remember, structural recursive
unificaton has been the top entry of my list of those funny bracketed,
painted programs that i would like to see in epigram.
Best regards,
Sebastian
