Re: [stack] An Overlooked Paradigm in Functional Programming

On Jan 30, 2010, at 6:13 PM, John Nowak wrote:

> Or, alternatively, I'd rather write:
>
>    fact = [zero?] [pop 1] [[] [1 -] bi fact *] ifte

For what it's worth, this is essentially identical to the FP version:

    fact = zero? -> ~1; * . <id, fact . pred>

The only real difference is that constant (~) is used instead of pop  
and that it's made explicit that 'fact' is applied only to the result  
of 'pred'. The latter is an advantage, although it's one that goes  
away when the stack-based version is written out visually (which is  
something I'll be harassing you all with soon enough).

You can do the same thing in Haskell of course:

    bi f g x = (f x, g x)
    ifte c f g x = if c x then f x else g x
    pred = flip (-) 1
    fact = ifte ((==) 0) (const 1) (uncurry (*) . bi id (fact . pred))

... but it's not amazingly pretty.

- jn