This is the mail archive of the
kawa@sources.redhat.com
mailing list for the Kawa project.
Re: bug in set! syntax
- From: Per Bothner <per at bothner dot com>
- To: Vladimir Tsichevski <wowa at jet dot msk dot su>
- Cc: Kawa List <kawa at sources dot redhat dot com>
- Date: Tue, 01 Apr 2003 10:53:18 -0800
- Subject: Re: bug in set! syntax
- References: <3E896D51.5090108@jet.msk.su>
Vladimir Tsichevski wrote:
Hi,
found a bug in set! sysntax handler. An attempt to use a syntax defined
as follows causes
"first set! argument is not a variable name" error
(define-syntax test
(syntax-rules ()
((test)
(let((s 1))
(set! s 1)
))))
(test)
Here is a patch that cures this very example,
Thanks! It's not quite right, since we want to
actually use the Symbol itself as the "name", not
just its String value.
I've checked in a modified patch allow with your test case.
but I'm afraid, there are
dosens of similar String/Symbol bugs in kawa code.
I think you're being a bit pessimistic - unless you've
examined the code, and found other places to fix?
This and similar problems are due to changes I made
last November to support use of Symbols as identifiers
as well as Strings. This generalization is needed (among
other reasons) if we're ever to support Common Lisp better.
--
--Per Bothner
per at bothner dot com http://per.bothner.com/