[plt-scheme] pattern matching and syntaxes

When I joined Indiana in 1984, I wrote all the EoPL/Hitchhiker 
interpreters in Prolog. Then I realized that I never needed 
backtracking; my life was simple because of unification, indeed, 
matching. I used Eugene's extend-syntax to port my Prolog matcher to 
Scheme and wrote all of Dan's assignments with it. It never occurred to 
me at the time that my matcher was in some way related to 
extend-syntax's.

Bruce and I continued to improve the matcher with only one thing in 
mind: making interpreters easy.

When Bruce joined me at Rice, we asked Andrew to check into the 
efficiency of matching because the MLers were writing a bunch of papers 
on their matchers. Andrew ported this technology and expanded the 
syntax further.

You probably know the recent history.

Before you try something different, explore the pragmatics! -- Matthias




On May 22, 2005, at 12:03 PM, Corey Sweeney wrote:

>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-scheme
>
> I'm considering writing a functional pattern matcher (as opposed to
> one implemented as a macro).   If I implement it, I might as well make
> it as as syntactically similar to one of the existing matchers, and I
> was wondering if anyone had reasons for/against going with a
> "define-syntax" like syntax over a "match" like syntax.
>
> Corey
>