[erlang-questions] View patterns

[Ulf Wiger (TN/EAB)]

ok wrote:
> 
> One of the things which is about to be added to GHC
> is view patterns.
> The idea is that if E is an expression and P is a
> pattern, then
> 	(E -> P)
> is a pattern.  This pattern matches a value V if
> and only if P = (E)(V).
> 
[...]
> 
> People who dislike n+k patterns in Haskell can no longer hurt 
> us much by taking them out.  With the aid of
> 
> 	np(K, X) when integer(X), integer(K), X >= K ->
> 	    {X};
> 	np(_, _) ->
> 	    false.
> 
> we can rewrite
> 
> 	factorial(0) -> 1;
> 	factorial(N+1) -> (N+1)*factorial(N).
> 
> as
> 
> 	factorial(0) -> 1;
> 	factorial((np(1) -> {N})) -> (N+1)*factorial(N).

Interesting stuff.

I'm a bit unsure about the semantics of this
last one. I wrote a parse transform to allow for 
experimentation with view patterns without revising
the Erlang grammar, and changed the above to:

   factorial(0) -> 1;
   factorial(_:_(np, [1], {N}) ->
     (N+1)*factorial(N).

(Obviously not an improvement over your syntax)

which was transformed into:

   np(K, X) when integer(X),
                 integer(K), X >= K ->
       {X};
   np(_, _) ->
       false.

   factorial(0) ->
       1;
   factorial(V73) ->
       case np(1, V73) of
           {N} ->
               (N + 1) * factorial(N);
           _ ->
               erlang:error(function_clause)
       end.

But something is obviously wrong, since this
will not terminate.

Could you please enlighten me?

BR,
Ulf W