Re: SCM_VALIDATE_...

[Jim Blandy <jimb at savonarola dot red-bean dot com>]

I'm coming in rather late on this discussion, so I may be missing the
point, but:

Within the interpreter, it's useful to have predicates that ask about
a number's *representation*, not its *value*.  The Scheme-level
predicates only ask about numbers' values, never their
representations.

So we could have a set of predicates that ask about representations:

  SCM_REP_FIXNUM_P
  SCM_REP_FLOAT_P
  SCM_REP_BIGNUM_P
  ...

and a set of predicates that ask about values (as the Scheme functions do):

  SCM_INTEGER_P
  SCM_RATIONAL_P
  SCM_REAL_P
  SCM_COMPLEX_P
  ...

  SCM_EXACT_P
  SCM_INEXACT_P

In general, I think user code really wants the latter set of
functions.  We should have a value-oriented set of conversion
functions, named after the C type we're converting from or to, but
which treat all representations similarly:

	scm_num2int
        scm_num2long
        scm_num2ulong
        scm_num2double
        ...

        scm_int2num
        scm_long2num
        scm_ulong2num
	scm_double2num
        ...

The idea here is that the conversions succeed if a "true" conversion
is possible:
- conversions from Scheme objects to C integers work on any representation,
  but fail if the value is non-integral, inexact, or doesn't fit
- conversions from C integers to Scheme numbers produce exact values,
  using whatever representation is appropriate

I think this is more consistent with the way the rest of Scheme works.