RE: [clisp-list] Problem with complex backquote expression

[Cross, M. <mc...@ir...>]

My message to the list was delayed because I sent it out as HTML.
Anyways, I spent more time looking at this, and sent a patch to Bruno
who accepted it and checked it into CVS.

I also found that this was already logged as bug 1308444.  My patch is
listed in the bug report:

http://sourceforge.net/tracker/?group_id=3D1355&atid=3D101355&func=3Ddeta=
il&ai
d=3D1308444=20



--
Matt Cross
Senior Lead Software Engineer
iRobot Corporation
63 South Avenue, Burlington, MA 01803
781-418-3373 (ph)
781-345-0201 (fax)
mc...@ir...
www.irobot.com <http://www.irobot.com/>=20
=20
=20


________________________________

	From: cli...@li...
[mailto:cli...@li...] On Behalf Of Cross,
Matthew
	Sent: Monday, March 06, 2006 9:47 AM
	To: cli...@li...
	Subject: [clisp-list] Problem with complex backquote expression
=09
=09
	(I tried to send this to the list on Friday, but it seems to
have gotten dropped)
	=20
	When I run this code with the latest release (2.38), I get
strange results:
	=20
	----
	(format t "~%~%test case 1:~%")
	=20
	(let* ((testme nil)
	       (testform ``(a ,@',testme d)) )
	  (princ testform)
	  (format t "~%")
	  (princ (eval testform)) )
	=20
	=20
	=20
	(format t "~%test case 2:~%")
	(let* ((testme '(b c))
	       (testform ``(a ,@',testme d)) )
	  (princ testform)
	  (format t "~%")
	  (princ (eval testform)) )
	=20
	=20
	=20
	(format t "~%~%")
	----
	=20
	I would expect evaluating testform to give me (a d) in the first
case and (a b c d) in the second case.  However, clisp seems to truncate
the list after the ,@ portion of the expression, giving me (a) and (a b
c) from the test cases.
	=20
	I narrowed it down to the problem being introduced in 2.33.84
(CVS tag clisp_2_33_84-2005-07-07).  When I run it in this release (and
the CVS head), I get:
	=20
	----
	[1]> (lisp-implementation-version)
	"2.33.84 (2005-07-07) (built 3350413236) (memory 3350413776)"
	[2]> (load "~/backquote-test.lisp")
	;; Loading file /home/mcross/backquote-test.lisp ...
	=20
	test case 1:
	'(A)
	(A)
	test case 2:
	'(A B C)
	(A B C)
	=20
	;; Loaded file /home/mcross/backquote-test.lisp
	T
	----
	=20
	When I run it in 2.33.2 (CVS tag clisp_2_33_2-2004-06-02) and
older releases, it works:
	=20
	----
	[1]> (lisp-implementation-version)
	"2.33.2 (2004-06-02) (built 3350412156) (memory 3350412304)"
	[2]> (load "~/backquote-test.lisp")
	;; Loading file /home/mcross/backquote-test.lisp ...
	=20
	test case 1:
	(CONS 'A (APPEND 'NIL '(D)))
	(A D)
	test case 2:
	(CONS 'A (APPEND '(B C) '(D)))
	(A B C D)
	=20
	;; Loaded file /home/mcross/backquote-test.lisp
	T
	----
	=20
	I am running on Ubuntu Linux 5.10.
	=20
	I don't see anything obvious in the NEWS file for 2.38 that
would affect this.  I'm going to look around to see if I can understand
what's going on, but I've never delved into clisp internals before.  Can
someone take a look at this also?
	=20
	Thanks,
	=20
	--
	Matt Cross
	Senior Lead Software Engineer
	iRobot Corporation
	63 South Avenue, Burlington, MA 01803
	781-418-3373 (ph)
	781-345-0201 (fax)
	mc...@ir...
	www.irobot.com <http://www.irobot.com/>=20
	=20