gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Package test failures


From: Paul F. Dietz
Subject: [Gcl-devel] Package test failures
Date: Fri, 11 Oct 2002 23:29:01 -0500
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.9) Gecko/20020408

I've classified the causes of the failures of the package tests:

EXPORT-4
   Export needs to signal a PACKAGE-ERROR when the symbol
   is not accessible in the package.  This error should be correctable.

MAKE-PACKAGE-ERROR-[1234]
DEFPACKAGE-24,25
   COMPUTE-RESTARTS needs to take an optional argument.

WITH-PACKAGE-ITERATOR-*
   The macro WITH-PACKAGE-ITERATOR isn't defined yet.

UNEXPORT-[56]
   UNEXPORT needs to throw a PACKAGE-ERROR on errors,
   and it should not throw an error if called on a symbol
   that is only internal in the package.

IN-PACKAGE-2
   The IN-PACKAGE macro should not evaluate its argument.

IN-PACKAGE-5
   For some reason the HANDLER-CASE is not catching the
   package error.

DEFPACKAGE-3
   Documentation is not being stored so that (DOCUMENTATION <package> T)
   can retrieve it.

DEFPACKAGE-4, DO-SYMBOLS-1,4,5
   These appear to be broken because DO-SYMBOLS is not correct.
   It needs to iterate over all the symbols accessible in a package,
   not just those present in the package.  Accessible symbols include
   those inherited from other packages.  Here, the package H has no
   internal or external symbols but does inherit the symbol A:FOO.

DEFPACKAGE-5,6,7,8,9,10,11,12
   These fail because a package is being made to have LISP on its
   use list.  The presence of the (:use) argument to DEFPACKAGE
   should have caused the package to have an empty use list.

DEFPACKAGE-13
   Repeating the :SIZE parameter should cause a PROGRAM-ERROR.

DEFPACKAGE-14
   Repeating the :DOCUMENTATION parameter should cause a PROGRAM-ERROR.

DEFPACKAGE-15,16
   Should signal a PACKAGE-ERROR if a nickname is already in use.

DEFPACKAGE-17,18,19,20,21,22,23
   These are name conflicts that should cause PROGRAM-ERRORs.


        Paul





reply via email to

[Prev in Thread] Current Thread [Next in Thread]