benefits of SRE syntax Michael Montague (16 Oct 2013 18:44 UTC)
Re: benefits of SRE syntax Roderic Morris (16 Oct 2013 19:23 UTC)
Re: benefits of SRE syntax Alex Shinn (20 Oct 2013 07:13 UTC)
Re: benefits of SRE syntax Per Bothner (16 Oct 2013 19:34 UTC)
Re: benefits of SRE syntax Alex Shinn (20 Oct 2013 14:21 UTC)
Re: benefits of SRE syntax John Cowan (20 Oct 2013 16:30 UTC)
Re: benefits of SRE syntax Per Bothner (20 Oct 2013 17:16 UTC)
Re: benefits of SRE syntax John Cowan (20 Oct 2013 17:50 UTC)
Re: benefits of SRE syntax Alex Shinn (20 Oct 2013 21:17 UTC)
Re: benefits of SRE syntax John David Stone (16 Oct 2013 20:39 UTC)
Re: benefits of SRE syntax Peter Bex (16 Oct 2013 20:50 UTC)
Re: benefits of SRE syntax Alex Shinn (17 Oct 2013 08:41 UTC)

Re: benefits of SRE syntax John Cowan 20 Oct 2013 17:50 UTC

Per Bothner scripsit:

> If "structured regular expressions" are to be part of the language,
> we should think about how they apply to sequences (lists and vectors)
> in general, not just strings.

I agree with that, and I think it makes sense to have an API for matching
SREs against general vectors and/or lists.  I don't believe, however,
that it should be integrated with this API.  Scheme is almost entirely
a monomorphic language, except for the exact/inexact polymorphism of the
numeric procedures.  (Of course, many procedures are either universally
polymorphic, like `cons`, or effectively universal, like `write`, which
is the union of a group of monomorphic procedures.)

Layering an object system on top of Scheme is a very reasonable thing
to do.  Constructing Scheme on top of such an existing system, as Kawa
does, is also reasonable.  Keeping the standard API monomorphic allows
either of these strategies or both at once.

--
Cash registers don't really add and subtract;           John Cowan
        they only grind their gears.                    xxxxxx@ccil.org
But then they don't really grind their gears, either;
        they only obey the laws of physics.  --Unknown