Re: Nil is a sequence.

[Jeff Dalton <jeff%aiva.edinburgh.ac.uk@nss.cs.ucl.ac.uk>]

> Date: Sun, 20 Dec 87 01:31:43 -0500
> From: quiroz@edu.rochester.cs

>       >#1a()
> 
>       Error: NIL is not a sequence.
>       Error signalled by SYSTEM:SHARP-A-READER.

> I am reasonable sure (upon careful exegesis of p.357) that my typo
> should have produced a vector of size zero.  [If not persuaded,
> consider (make-array 0 :initial-contents ())]

I agree.  The code in SI:SHARP-A-READER takes one more (elt ic 0)
than it needs to but gets away with it only if ic *has* a 0th elt.

Here is a (somewhat messy) fix:

(defun si:sharp-a-reader (stream subchar arg)
  (declare (ignore subchar))
  (let ((initial-contents (read stream nil nil t)))
    (if *read-suppress*
        nil
        (let ((ic initial-contents)
	      (d nil))
	  (do ((i 0 (1+ i)))
	      ((>= i arg)
	       (make-array (nreverse d) :initial-contents initial-contents))
	    (push (length (if (> i 0)
			      (setq ic (elt ic 0))
			      ic))
		  d))))))

Unfortunately, a similar problem will still appear in MAKE-ARRAY
if you type something like #2a(()()).  Or try

     (make-array '(2 0) :initial-contents '(()()))


> Even if this interpretation is not right, the error message lies.

This is ELT's fault.  You'll get the same error if you try (elt () 0).

Jeff Dalton,                      JANET: J.Dalton@uk.ac.ed             
AI Applications Institute,        ARPA:  J.Dalton%uk.ac.ed@nss.cs.ucl.ac.uk
Edinburgh University.             UUCP:  ...!ukc!ed.ac.uk!J.Dalton