This is the mail archive of the
kawa@sourceware.org
mailing list for the Kawa project.
RE: extended "colon notation" [feedback needed]
- From: <Fernando dot Matomira at thomson dot com>
- To: <per at bothner dot com>
- Cc: <kawa at sources dot redhat dot com>
- Date: Thu, 4 Aug 2005 09:45:32 +0200
- Subject: RE: extended "colon notation" [feedback needed]
I was arguing some years ago in comp.lang.lisp for the need of
hierarchical packages in Common Lisp, eg. foo:bar:baz , foo:bar::baz
(which currently gives an error in CL). As this is a compatible
extension, maybe you could be the first one to implement it, so the idea
gets pushed over there.
-----Original Message-----
From: Per Bothner [mailto:per@bothner.com]
Sent: Thursday, August 04, 2005 9:33 AM
To: Mato Mira, Fernando (DIA EMEA/A)
Cc: kawa@sources.redhat.com
Subject: Re: extended "colon notation" [feedback needed]
Fernando.Matomira@thomson.com wrote:
> Does this cause any problems to people using symbols with colons
> in them for compatibility with Common Lisp?
> For example, CL:IF macro.
I don't think so. In fact, Kawa already implements Common Lisp-style
"two-level" symbols: The Symbol class points to a Namespace (which is
more-or-less a Common Lisp package) plus a "local name". The Symbol
class isn't normally used for Scheme symbols, but there is a good chance
that will change.
One difference is that Kawa supports an extra level of indirection
between "prefixes" and namespaces, and the prefixes may be lexically
scoped. This is needed/derirable for XML processing, and the XQuery
language. I.e. "package nicknames" may be local rather than global.
The goal is to support the full Common Lisp semantics for symbols and
packages, though it's not a very high priority.
--
--Per Bothner
per@bothner.com http://per.bothner.com/