[ANN] R7RS Scheme
Kjetil Svalastog Matheussen
R7RS Scheme Draft V2007-4-1
***************************
***************************
About
=====
R7RS Scheme is a Lisp dialect largely inspired by Scheme.
Summary
=======
R7RS Scheme is a statically scoped and properly tail-recursive dialect of
the Lisp programming language. It was designed to have an exceptionally
clear and simple semantics and few different ways to form expressions. A
wide variety of programming paradigms, including imperative, functional,
and message passing styles, find convenient expression in R7RS Scheme.
Introduction
============
Programming languages should be designed not by piling feature on top of
feature, but by removing the weaknesses and restrictions that make
additional features appear necessary. R7RS Scheme demonstrates that a very
small number of rules for forming expressions, with no restrictions on how
they are composed, suffice to form a practical and efficient programming
language that is flexible enough to support most of the major programming
paradigms in use today.
Implementation
==============
http://www.notam02.no/~kjetism/r7rs.tar.gz
Changes between R7RS Scheme and Scheme[1]
=========================================
* "define" can be used anywhere in an expression, just like on the top-level:
(let ((a 1))
(set! a 2)
(define a 3)
a)
=> 3
(let ((a 1))
(set! a 2)
(define (a) (c))
(define (c) 3)
(a))
=> 3
(let ((a 1))
(set! a 2)
(define (a) (c))
(define (c) 3)
(define a (+ (a) 1))
a)
=> 4
* "include" includes code from another file:
(define (cdr-list-copy list)
(include "srfi-1.scm")
(cdr (list-copy list)))
* "define-toplevel":
(let ((a* 1))
(define-toplevel (a)
a*))
(a)
=> 1
(let ((a* 1)
(funcname 'a))
(define-toplevel (,funcname)
a*))
(a)
=> 1
* "unquote" and "unquote-splicing" can appear outside quasiquotes:
(let ((a 1))
,'a)
=> 1
(let ()
,(+ 8 5))
=> 14
("(+ 8 6)" was calculated at compile time, and not at run time)
(let ((a 1)
(b 2))
,@(map (lambda (val)
`(begin
(display (string-append (symbol->string ',val) "=" (number->string ,val)))
(newline)))
'(a b)))
=> "a=1"
"b=2"
* Arguments are evaluated in order left to right and top to bottom:
(let ((a 1))
(list (begin
(set! a 2)
3)
a))
=> (3 2)
(let ((a 1))
`(,(begin
(set! a 2)
3)
,a))
=> (3 2)
(let ((a 1))
(let ((b (begin
(set! a 2)
3))
(c a))
(list b c)))
=> (3 2)
* "map" works like "map-in-order" as defined in scheme/srfi-1.
* Calling "car" on the empty list returns the empty list:
(car '())
=> ()
* Calling "cdr" on the empty list returns the empty list:
(cdr '())
=> ()
* "eval" takes one argument only:
(let ((a 1))
(eval '(+ a 2)))
=> 3
* "letrec*":
(letrec* ((a d)
(b (lambda () (c)))
(c (lambda () a))
(d 1))
(b))
=> 1
* Symbols, keywords, syntax and function names are case sensitive.
* "define-macro" does not need to be placed at the top-level.
However, the following expression does not work:
(let ((a 1))
(define-macro (gakk val)
`(+ ,val a)))
(gakk 2)
=> Error, undefined varible "a" in expression "(gakk 2) => (+ 2 a)"
* "macroexpand" and "macroexpand-1" are guaranteed to return expressions
where the transformations are defined only by the use of "define-macro".
For example, "define-macro" is not used to implement core functionality
such as "cond" or "letrec*". So this will always work in a clean
environment:
(macroexpand '(cond (#t 5)(else 6)))
=> (cond (#t 5)(else 6))
* "(gensym)" returns a unique symbol. Handy for macros.
* Macros can be used as functions:
(define-macro (add . rest)
`(+ ,@rest))
(apply add '(1 2))
=> 3
(let ((list '(1 2)))
(apply add list)
=> 3
(map add '(1 2) '(3 4))
=> '(4 6)
Max 30 arguments, and the macro must be defined before being referenced.
* Keywords:
(list :this-is-a-keyword 'this-is-a-symbol)
=> (:this-is-a-keyword this-is-a-symbol)
(keyword->symbol :gakk)
=> gakk
(symbol->keyword 'gakk)
=> :gakk
(keyword->string :gakk)
=> "gakk"
(string->keyword "gakk")
=> :gakk
(keyword? :gakk)
=> #t
(keyword? 'gakk)
=> #f
* No hygenic macros. (Programming languages should be designed not by
piling feature on top of feature, but by removing the weaknesses and
restrictions that make additional features appear necessary).
Notes
=====
[1] As defined by the r5rs specification.
William D Clinger
> R7RS Scheme is a Lisp dialect largely inspired by Scheme.
It's about time for Scheme to inspire a dialect of Lisp.
Will
Pascal Bourguignon
> R7RS Scheme Draft V2007-4-1
Sounds good. Perhaps R9RS will be indistinguishable enough from CL
that I'll be able to use it :-)
--
__Pascal Bourguignon__
http://www.informatimago.com
http://pjb.ogamita.org
Didier Verna
> Kjetil Svalastog Matheussen <kje...@notam02.no> writes:
>> R7RS Scheme Draft V2007-4-1
>
> Sounds good. Perhaps R9RS will be indistinguishable enough from CL
> that I'll be able to use it :-)
Actually, no way ! Quoting Mike Sperber at that Thai restaurant just 1
hour ago: "R6RS is just, you know, flushing the 80's. Objects are the 90's.
We're not quite there yet..." :-)
I think I'll put that one in my .sig for a while :-)
--
Didier Verna, did...@lrde.epita.fr, http://www.lrde.epita.fr/~didier
EPITA / LRDE, 14-16 rue Voltaire Tel.+33 (1) 44 08 01 85
94276 Le Kremlin-Bicętre, France Fax.+33 (1) 53 14 59 22 did...@xemacs.org
David Rush
> Quoting Mike Sperber at that Thai restaurant just 1
> hour ago: "R6RS is just, you know, flushing the 80's. Objects are the 90's.
> We're not quite there yet..." :-)
Oh no. Please tell me he was just talking shite after drinking too
much Asian beer. I thought that the category-theory people had all
made this abundantly clear during the late 90s ;)
--
The Scheme Underground deals harshly with recidivists. Have your self-
criticism prepared in advance.
Michael Sperber
"David Rush" <kumo...@gmail.com> writes:
> On Apr 2, 10:18 pm, Didier Verna <did...@lrde.epita.fr> wrote:
>> Quoting Mike Sperber at that Thai restaurant just 1
>> hour ago: "R6RS is just, you know, flushing the 80's. Objects are the 90's.
>> We're not quite there yet..." :-)
>
> Oh no. Please tell me he was just talking shite after drinking too
> much Asian beer. I thought that the category-theory people had all
> made this abundantly clear during the late 90s ;)
Be that as it may, but you don't have worry, because ...
I WAS SPEAKING AS AN INDIVIDUAL MEMBER OF THE SCHEME COMMUNITY. I WAS
NOT SPEAKING FOR THE R6RS EDITORS, AND THIS QUOTE SHOULD NOT BE
CONFUSED WITH THE EDITORS' OFFICIAL POSITION.
:-)
--
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla
Matthias Blume
> On Apr 2, 10:18 pm, Didier Verna <did...@lrde.epita.fr> wrote:
>> Quoting Mike Sperber at that Thai restaurant just 1
>> hour ago: "R6RS is just, you know, flushing the 80's. Objects are the 90's.
>> We're not quite there yet..." :-)
>
> Oh no. Please tell me he was just talking shite after drinking too
> much Asian beer. I thought that the category-theory people had all
> made this abundantly clear during the late 90s ;)
Not to worry. We are well into the 00's. Objects are soooo last
millenium...
:-)
William D Clinger
> I WAS SPEAKING AS AN INDIVIDUAL MEMBER OF THE SCHEME COMMUNITY. I WAS
> NOT SPEAKING FOR THE R6RS EDITORS, AND THIS QUOTE SHOULD NOT BE
> CONFUSED WITH THE EDITORS' OFFICIAL POSITION.
>
> :-)
Indeed, the R6RS editors don't have any official position on
the decade in which they are living.
Will
Didier Verna
> Be that as it may, but you don't have worry, because ...
>
> I WAS SPEAKING AS AN INDIVIDUAL MEMBER OF THE SCHEME COMMUNITY. I WAS
> NOT SPEAKING FOR THE R6RS EDITORS, AND THIS QUOTE SHOULD NOT BE
> CONFUSED WITH THE EDITORS' OFFICIAL POSITION.
>
> :-)
Right :-)
> "David Rush" <kumo...@gmail.com> writes:
>
>> On Apr 2, 10:18 pm, Didier Verna <did...@lrde.epita.fr> wrote:
>>> Quoting Mike Sperber at that Thai restaurant just 1
>>> hour ago: "R6RS is just, you know, flushing the 80's. Objects are the 90's.
>>> We're not quite there yet..." :-)
>>
>> Oh no. Please tell me he was just talking shite after drinking too
>> much Asian beer.
He was drinking ice tea AFAICR. *I* was drinking too much asian beer :-)