Re: define-syntax can only be used with local variables

[Per Bothner <per@bothner.com>]

On 9/17/21 12:00 AM, phiroc@free.fr wrote:
> Hi Per,
> 
> thank you for you input.
> 
> The R7RS report says the following:
> 
> "If the define-syntax occurs at the outermost level, then
> the global syntactic environment is extended by binding
> the keyword to the specified transformer, but previous
> expansions of any global binding for keyword remain unchanged.
> Otherwise, it is an internal syntax definition, and
> is local to the body in which it is defined. Any use of a
> syntax keyword before its corresponding definition is an
> error. In particular, any use that precedes an inner definition
> will not apply an outer definition."
> 
> So, as far as I am concerned, Kawa's behavior is the norm:
> a previously-defined global variable can't be reset by a define-syntax.

That is not what R7RS is saying.  It is talking about when you define a
keyword FOO using define-syntax at the outermost level; then use (expand) FOO;
then redefine FOO: In that case the previous expansions of FOO are unchanged.
This is in contrast to a function re-definition, where calls to FOO
in old calls should use the new FOO.

Kawa by default does some inlining that conflicts with this goal, hence the
kludge solution for issue #44.  Kawa has a --no-inline flag that makes Kawa
behave more like a traditional Scheme REPL.  Doing --no-inline has some
problems, but perhaps a modified --no-inline default for REPLs might work.

-- 
	--Per Bothner
per@bothner.com   http://per.bothner.com/