Re: [Chicken-users] Parsing lists from external scheme intepreters

[Michael Erdmann]

felix wrote:
> Michael Erdmann wrote:
>
> > Now the complicated part starts. FramerD was providing the concept
> > of choices in order to represent the contents of result sets.
> > The nice thing about it was the following trick:
> >
> > (define x (choice 1 2 3 4 5) )
> >
> > (define (inc y)  (+ y 1))
> >
> > (inc x) -> (choice 2 3 4 5 6)
> >
> > which means the operation inc was executed on any member in the
> > choice. I have not seen this by now in the chicken documentation.
> > Does some thing similar exisit some where?
>
>
> No, chicken doesn't support anything like that (and no other Scheme
> system I know, AFAIK). Depending on how desperately you need that,
> once could hack around with error-handlers, etc. But I don't
> recommend that, unless you *really* need it.
Thanks, but i will try to do a WoA. Choices are realy nice but
they are also very in secure, e.g:

      (write "Hallo") ...
      (print-hallo names)
      (write "Done.")

What happens, that print-hallo is never called if names is empty.
The result is a completly different trace fingerprint which does
not contain the print-hallo procedure at all. This is very confusing.
> > I guess i can represent a choice as a list of elements where
> > the list is somehow marked as a choice.
>
>
> Or try records, like SRFI-9 (`define-record-type'). That makes
> your choice object distinct from other data
>
> cheers,
> felix