Re: TYPES doesn't test COMPLEX

[Taiichi Yuasa <yuasa%tutics.tut.junet%utokyo-relay.csnet@RELAY.CS.NET>]

	Date: Thu, 10 Dec 87 16:51:03 PST
	From: raible%rodan.nas.nasa.gov%ngp.utexas.edu@RELAY.CS.NET L. Raible" <raible@rodan.nas.nasa.gov>
	Message-Id: <8712110051.AA01105@rodan.nas.nasa.gov>
	To: jeff%aiva.edinburgh.ac.uk@nss.cs.ucl.ac.uk
	Cc: @sally.utexas.edu
	In-Reply-To: Jeff Dalton's message of Thu, 10 Dec 87 23:50:10 GMT <14144.8712102350@aiva.ed.ac.uk>
	Subject: TYPEP doesn't test COMPLEX
	Received: from CSNet-Relay by utokyo-relay; 11 Dec 87 12:41:32-JST (Fri)
	Status: R


	   From: Jeff Dalton <jeff%aiva.edinburgh.ac.uk@nss.cs.ucl.ac.uk>
	   Date: Thu, 10 Dec 87 23:50:10 GMT

	   (TYPEP x 'COMPLEX) always => NIL.

		KCl (Kyoto Common Lisp)  June 3, 1987

		>(typep #c(1 1) 'complex)
		NIL

	   This can be fixed as follows:

		(setf (get 'complex 'type-predicate) 'complexp)

	   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

	of course this should have been

		(setf (get 'complex 'si::type-predicate) 'complexp)

	for people who live in the USER package.


There was a bug in the definition of the Lisp function TYPEP
in lsp/predlib.lsp.  To fix the bug, replace the CASE clause

    (complex
     (and (complexp object)
          (typep (realpart object) (car i))
          (typep (imagpart object) (car i))))
with
    (complex
     (and (complexp object)
          (or (null i)
              (and (typep (realpart object) (car i))
                   (typep (imagpart object) (car i))))))
-- Taiichi