Re: Odd Behavior

[Andy Wingo <wingo@pobox.com>]

On Mon 31 Jan 2011 19:44, Noah Lavine <noah.b.lavine@gmail.com> writes:

> (define (codegen)
>   #`(let ((message "hello, world!\n"))
>       (display message)))
>
> (define-syntax test
>   (lambda (x)
>     (syntax-case x ()
>       (_ (codegen)))))
>
> (test)
>
> Guile tried to compile test.scm, failed, and then interpreted it
> correctly. Is that right?

Yep; as mentioned in the NEWS entry, "** Functions needed by macros at
expand-time need to be present at expand-time."  I'm sure you know that
already, and your real question:

> is there a way to get more information about where the wrong-type-arg
> error was thrown?

There are several ways this could be improved:

  * The exception should be about an unbound variable, not about
    applying #f.  I'll see what I can do there.

  * The exception could be printed with an appropriate printer.  See
    http://thread.gmane.org/gmane.lisp.guile.devel/11066/focus=11158.

  * You should see the source location of the error and the containing
    forms.

Probably that thread is the best bet for making a start at this.

Regards,

Andy
-- 
http://wingolog.org/