Re: No gh_set_x()?

[Sam Tregar]

On Tue, 10 Jul 2001, Brett Viren wrote:

> I guess to get the equiv of set! you need to check if the symbol
> exists before using gh_define().  But, I am no expert.  Below is the
> reply I got when I asked this.  Hope it helps.

Ok, I'll respond to that below.  Maybe the original answerer can shed more
light on the subject.

>  > `set!' is not a procedure.  it's an "overloaded" syntax for two
>  > different things:
>  >
>  > * changing the binding of a lexical variable.  this is impossible to
>  >   express in C.

Um, why not?  It seems to me that given an SCM I should be able to set it
to the value given by another SCM and have that new value reflected
through whatever was pointing the the first SCM.

Perhaps there's a better way to ask for what I need - how would you
implement a "+=" procedure in C?  It takes two arguments, adds them and
modifies the initial argument to contain the result:

>  > * changing the binding of a top-level variable.  this would be
>  >   possible to do if the concept of a top-level variable (or a module)
>  >   was somehow exported in the gh_ interface, but it is not.
>  >
>  > so I guess that you want to achieve the effect of `set!'ing a
>  > top-level variable.  `gh_define' should do the trick, I believe.

Hm, ok.  The problem here is that by the time I get to the point where I'm
trying to call the non-existent gh_set_x() all I have is a single SCM.
How can I know if it's bound at the top-level or not?

-sam