guile-user
[Top][All Lists]
Advanced

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

Re: Re: intern a top level variable


From: tomas
Subject: Re: Re: intern a top level variable
Date: Wed, 16 Aug 2023 19:58:54 +0200

On Wed, Aug 16, 2023 at 06:13:05PM +0200, Jean Abou Samra wrote:
> Le mercredi 16 août 2023 à 10:55 -0400, Mortimer Cladwell a écrit :
> > I would like to intern and assign a value within a method:
> > 
> > (define (test-intern)
> > (let* ((name "abc")
> >        (data "def")
> >        (name-symbol (gensym name))
> >        )
> >   (pretty-print (string-append "symbol: " (symbol->string name-symbol)))
> >   (set! name-symbol data)))
> > 
> > scheme@(guile-user)> (test-intern)
> > "symbol: abc3301"
> > scheme@(guile-user)> abc3301
> > ;;; <unknown-location>: warning: possibly unbound variable `abc3301'
> > ERROR: In procedure module-lookup: Unbound variable: abc3301
> 
> 
> Sorry, but it's not clear to me what you mean by "intern and assign a value",
> and I don't think it will be clear to someone else.
> 
> Can you be more precise please?

I think what the OP wants is to make a binding for
an existing symbol. Something, perhaps like

  scheme@(guile-user)> (define foovar (gensym "foo"))
  scheme@(guile-user)> foovar
  $1 = foo170
  scheme@(guile-user)> (module-add! (current-module) foovar (make-variable 42))
  scheme@(guile-user)> foo170
  $2 = 42

(note that the "top level" is Just Another Module).

Mortimer: is that what you were looking for? Note that
"you usually don't do this" (whatever that means), so
it'd make sense to take a step back and think about what
you are trying to achieve :-)

> I don't understand why you expect "abc3301" to be bound after running (test-
> intern). That (test-intern) call just creates a symbol, which is interned as a
> symbol, meaning that creating another (interned) symbol with the same name 
> will
> reuse the same symbol value. But there is no reason why it shoud be bound to a
> variable. Symbol interning just applies to symbols as values, it has nothing 
> to
> do with variables.

In any case it would make sense to revisit the concepts
of "symbol", "variable" and "binding". It's subtle, but
in Lispy lands it is indispensable.

Cheers
-- 
t

Attachment: signature.asc
Description: PGP signature


reply via email to

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