Re: bug in psyntax?

[email protected] Fri, 2 May 2003 13:27:57 -0500

Newsgroups	gmane.comp.java.sisc.devel
Message-ID	<[email protected]>

On Fri, May 02, 2003 at 06:56:45PM +0100, Matthias Radestock wrote:
> (let ((x 1))
>   (let-syntax ([foo (syntax-rules () ((_) 2))])
>     (define x (foo))
>     3)
>   x)
> 
> => 2
> 
> (sc-expand '(let ((x 1))
>               (let-syntax ([foo (syntax-rules () ((_) 2))])
>                 (define x (foo))
>                 3)
>               x))
> 
> =>
> 
> ((lambda (|x_GAX0qWu6h|)
>    (letrec ((|x_GAFKfbx6h| '2))
>      (begin '3 |x_GAFKfbx6h|)))
>  '1)
> 
> It appears that let-syntax is not creating a local scope, letting the 
> binding for x escape into the surrounding scope.
> 
> Here is what R5RS has to say:
> 
> [section 5.2.2]
> Definitions may occur at the beginning of a <body> (that is, the body of 
> a lambda, let, let*, letrec, let-syntax, or letrec-syntax expression or 
> that of a definition of an appropriate form). [...]
> The variable defined by an internal definition is local to the <body>.
> 
> 
> To me this pretty unambiguously requires the above example to return 1, 
> not 2.
> 
> Have I missed something, or is this a bug in psyntax?

Looks like a bug to me, forwarded to Dybvig.

	Scott

signature.asc (application/pgp-signature, 189 B)

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+sritr9IW4v3mHtQRAgV8AJ9krPUixoTmy1L/Qqh4UOpoS1FqtwCeLfRa
TCKp8GimUyzMuMXlCxZx7tY=
=BMf3
-----END PGP SIGNATURE-----