Re: [Axiom-mail] A pattern matching question

[Martin Rubey]

Themos Tsikas <address@hidden> writes:

> --try this in an input file
> 
> tt:=operator 'tt
> ttrules := rule
>   tt(n^(a | integer?(a))) == t/a
>   tt(n/(n-1)) == 1/t
>   tt(n/(n^a+(X | integer?(a) and (n-1)^a = n^a +X))) == 1/t^a
> 
> ttrules tt(n^6)
> ttrules tt(n/(n-1))
> ttrules tt(n/(n-1)^2)
> ttrules tt(n/(n-1)^30)

Ui, I like that.  Strange that it can apply such a complicated rule but stumble
over freeOf?

Alasdair, just to make it completely obvious:  

(3) -> ttrules tt(n/(n-1)^2)

                 n
   (3)   tt(-----------)
             2
            n  - 2n + 1

is because n^2-2n+1 is different from (n-1)^a.  As far as I can see, the match
itself is tried against the pattern without trying to "expand" or transform it
in some way.  (Although I must admit that I didn't check.)

Martin