Subject: Re: empty list?
From: (Rob Warnock)
Date: 2000/05/16
Newsgroups: comp.lang.scheme
Message-ID: <8fqi08$bertm$>
John Perks  <> wrote:
| Is it true that Lisp's empty list, unlike Scheme's, has car and cdr equal
| to itself? This seems to be true of the Lisps I've tried it on, but what
| the Standard say?

As Barry said, for CL (c{a,d}r nil) ==> nil, but note that this does *not*
imply that the empty list is necessarily implemented as some distinguished
cons cell which has both its car & cdr pointers pointing to itself [despite
some early implementations which did exactly that]. In fact, in neither
Scheme nor CL is the empty list *permitted* to be a [user-level] cons cell:

	% clisp
	> (consp '())
	> ^D
	% mzscheme
	> (pair? '())
	> ^D

The difference is that Scheme requires the argument of "car" or "cdr"
to be a pair (cons cell), while CL requires the argument of "car" or "cdr"
to be *either* a cons cell or the empty list -- and in the latter case the
standard happens to *define* the value of the operation to be the empty list.

One could get the same behavior in Scheme, if you like, by redefining the
built-in "car" & "cdr" [purists, please, no flames!] as follows:

	> (car '())
	car: expects argument of type <pair>; given ()
	> (let ((original-car car))
	    (set! car
	      (lambda (x)
		  ((pair? x) (original-car x))
		  ((null? x) x)
		  (else (error "car: not a list:" x))))))
	> (car '())


Rob Warnock, 41L-955
Applied Networking
Silicon Graphics, Inc.		Phone: 650-933-1673
1600 Amphitheatre Pkwy.		PP-ASEL-IA
Mountain View, CA  94043