cees-bart <[EMAIL PROTECTED]> writes:
> (make-package :my-hash-utils)
MY-HASH-UTILS will use the COMMON-LISP -package by default.
> (in-package :my-hash-utils)
> (export 'make-hash-table)
This refers to COMMON-LISP:MAKE-HASH-TABLE.
> (defun make-hash-table (&key (test 'eql) (hash-function 'eql-hash)
> (size 65) (rehash-size 1.5)
> (rehash-threshold 1) (weak-p nil))
This redefines COMMON-LISP:MAKE-HASH-TABLE...
> (common-lisp:make-hash-table :test test
... so no doubt you get recursion. To fix this, shadow
make-hash-table in the my-hash-utils package too:
(defpackage "MY-HASH-UTILS"
(:use "COMMON-LISP")
(:shadow "MAKE-HASH-TABLE"))
Earlier, you wrote:
> I know that redefining is very dangerous, but note that even if I
> 'redefine' make-hash-table with the _same_ sourcecode, the stackoverflow
> occurs. Any clues?
Perhaps the interpreter uses MAKE-HASH-TABLE. It seems to work
if I compile it with (compile 'make-hash-table #'(lambda ...)) or
with compile-file.
-- Attached file included as plaintext by Listar --
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
iD8DBQBBkRuEHm9IGt60eMgRAt4LAKC1aEDvtPaxk4tN7IUyp+8Rl5yCGQCfZxpk
EZtBkwLJ5mMdDK93qcJfcZ4=
=If5z
-----END PGP SIGNATURE-----
