607 lines
24 KiB
HTML
607 lines
24 KiB
HTML
|
<!DOCTYPE HTML PUBLIC "-//W3O//DTD W3 HTML 2.0//EN">
|
||
|
|
<!Converted with LaTeX2HTML 0.6.5 (Tue Nov 15 1994) by Nikos Drakos (nikos@cbl.leeds.ac.uk), CBLU, University of Leeds >
|
||
|
|
<HEAD>
|
||
|
|
<TITLE>Appendix C. Backquote</TITLE>
|
||
|
|
</HEAD>
|
||
|
|
<BODY>
|
||
|
|
<meta name="description" value=" Backquote">
|
||
|
|
<meta name="keywords" value="clm">
|
||
|
|
<meta name="resource-type" value="document">
|
||
|
|
<meta name="distribution" value="global">
|
||
|
|
<P>
|
||
|
|
<b>Common Lisp the Language, 2nd Edition</b>
|
||
|
|
<BR> <HR><A NAME=tex2html6284 HREF="node368.html"><IMG ALIGN=BOTTOM ALT="next" SRC="icons/next_motif.gif"></A> <A NAME=tex2html6282 HREF="clm.html"><IMG ALIGN=BOTTOM ALT="up" SRC="icons/up_motif.gif"></A> <A NAME=tex2html6276 HREF="node366.html"><IMG ALIGN=BOTTOM ALT="previous" SRC="icons/previous_motif.gif"></A> <A NAME=tex2html6286 HREF="node1.html"><IMG ALIGN=BOTTOM ALT="contents" SRC="icons/contents_motif.gif"></A> <A NAME=tex2html6287 HREF="index.html"><IMG ALIGN=BOTTOM ALT="index" SRC="icons/index_motif.gif"></A> <BR>
|
||
|
|
<B> Next:</B> <A NAME=tex2html6285 HREF="node368.html">References</A>
|
||
|
|
<B>Up:</B> <A NAME=tex2html6283 HREF="clm.html">Common Lisp the Language</A>
|
||
|
|
<B> Previous:</B> <A NAME=tex2html6277 HREF="node366.html"> Discussion</A>
|
||
|
|
<HR> <P>
|
||
|
|
<H1><A NAME=SECTION003600000000000000000>Appendix C.<br>Backquote</A></H1>
|
||
|
|
<P>
|
||
|
|
<img align=bottom alt="change_begin" src="gif/change_begin.gif"><br>
|
||
|
|
<A NAME=BACKQUOTESIMULATOR>Here</A>
|
||
|
|
is the code for an implementation of backquote syntax
|
||
|
|
(see section <A HREF="node190.html#BACKQUOTE">22.1.3</A>) that I have found quite useful
|
||
|
|
in explaining to myself the behavior of nested backquotes.
|
||
|
|
It implements the formal rules for backquote processing
|
||
|
|
and optionally applies a code simplifier to the result.
|
||
|
|
One must be very careful in choosing the simplification rules;
|
||
|
|
the rules given here work, but some Common Lisp implementations
|
||
|
|
have run into trouble at one time or another by using a
|
||
|
|
simplification rule that does not work in all cases.
|
||
|
|
Code transformations that are plausible when single forms
|
||
|
|
are involved are likely to fail in the presence of splicing.
|
||
|
|
<P>
|
||
|
|
At the end of this appendix are some samples of
|
||
|
|
nested backquote syntax with commentary.
|
||
|
|
<P>
|
||
|
|
<P><pre>
|
||
|
|
;;; Common Lisp backquote implementation, written in Common Lisp.
|
||
|
|
;;; Author: Guy L. Steele Jr. Date: 27 December 1985
|
||
|
|
;;; Tested under Symbolics Common Lisp and Lucid Common Lisp.
|
||
|
|
;;; This software is in the public domain.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; $ is pseudo-backquote and % is pseudo-comma. This makes it
|
||
|
|
;;; possible to test this code without interfering with normal
|
||
|
|
;;; Common Lisp syntax.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; The following are unique tokens used during processing.
|
||
|
|
;;; They need not be symbols; they need not even be atoms.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defvar *comma* (make-symbol "COMMA"))
|
||
|
|
(defvar *comma-atsign* (make-symbol "COMMA-ATSIGN"))
|
||
|
|
(defvar *comma-dot* (make-symbol "COMMA-DOT"))
|
||
|
|
(defvar *bq-list* (make-symbol "BQ-LIST"))
|
||
|
|
(defvar *bq-append* (make-symbol "BQ-APPEND"))
|
||
|
|
(defvar *bq-list** (make-symbol "BQ-LIST*"))
|
||
|
|
(defvar *bq-nconc* (make-symbol "BQ-NCONC"))
|
||
|
|
(defvar *bq-clobberable* (make-symbol "BQ-CLOBBERABLE"))
|
||
|
|
(defvar *bq-quote* (make-symbol "BQ-QUOTE"))
|
||
|
|
(defvar *bq-quote-nil* (list *bq-quote* nil))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; Reader macro characters:
|
||
|
|
;;; $foo is read in as (BACKQUOTE foo)
|
||
|
|
;;; %foo is read in as (#:COMMA foo)
|
||
|
|
;;; %@foo is read in as (#:COMMA-ATSIGN foo)
|
||
|
|
;;; %.foo is read in as (#:COMMA-DOT foo)
|
||
|
|
;;; where #:COMMA is the value of the variable *COMMA*, etc.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; BACKQUOTE is an ordinary macro (not a read-macro) that
|
||
|
|
;;; processes the expression foo, looking for occurrences of
|
||
|
|
;;; #:COMMA, #:COMMA-ATSIGN, and #:COMMA-DOT. It constructs code
|
||
|
|
;;; in strict accordance with the rules on pages 349-350 of
|
||
|
|
;;; the first edition (pages 528-529 of this second edition).
|
||
|
|
;;; It then optionally applies a code simplifier.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(set-macro-character #\$
|
||
|
|
#'(lambda (stream char)
|
||
|
|
(declare (ignore char))
|
||
|
|
(list 'backquote (read stream t nil t))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(set-macro-character #\%
|
||
|
|
#'(lambda (stream char)
|
||
|
|
(declare (ignore char))
|
||
|
|
(case (peek-char nil stream t nil t)
|
||
|
|
(#\@ (read-char stream t nil t)
|
||
|
|
(list *comma-atsign* (read stream t nil t)))
|
||
|
|
(#\. (read-char stream t nil t)
|
||
|
|
(list *comma-dot* (read stream t nil t)))
|
||
|
|
(otherwise (list *comma* (read stream t nil t))))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
|
||
|
|
;;; If the value of *BQ-SIMPLIFY* is non-NIL, then BACKQUOTE
|
||
|
|
;;; processing applies the code simplifier. If the value is NIL,
|
||
|
|
;;; then the code resulting from BACKQUOTE is exactly that
|
||
|
|
;;; specified by the official rules.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defparameter *bq-simplify* t)
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defmacro backquote (x)
|
||
|
|
(bq-completely-process x))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; Backquote processing proceeds in three stages:
|
||
|
|
;;;
|
||
|
|
;;; (1) BQ-PROCESS applies the rules to remove occurrences of
|
||
|
|
;;; #:COMMA, #:COMMA-ATSIGN, and #:COMMA-DOT corresponding to
|
||
|
|
;;; this level of BACKQUOTE. (It also causes embedded calls to
|
||
|
|
;;; BACKQUOTE to be expanded so that nesting is properly handled.)
|
||
|
|
;;; Code is produced that is expressed in terms of functions
|
||
|
|
;;; #:BQ-LIST, #:BQ-APPEND, and #:BQ-CLOBBERABLE. This is done
|
||
|
|
;;; so that the simplifier will simplify only list construction
|
||
|
|
;;; functions actually generated by BACKQUOTE and will not involve
|
||
|
|
;;; any user code in the simplification. #:BQ-LIST means LIST,
|
||
|
|
;;; #:BQ-APPEND means APPEND, and #:BQ-CLOBBERABLE means IDENTITY
|
||
|
|
;;; but indicates places where "%." was used and where NCONC may
|
||
|
|
;;; therefore be introduced by the simplifier for efficiency.
|
||
|
|
;;;
|
||
|
|
;;; (2) BQ-SIMPLIFY, if used, rewrites the code produced by
|
||
|
|
;;; BQ-PROCESS to produce equivalent but faster code. The
|
||
|
|
;;; additional functions #:BQ-LIST* and #:BQ-NCONC may be
|
||
|
|
;;; introduced into the code.
|
||
|
|
;;;
|
||
|
|
;;; (3) BQ-REMOVE-TOKENS goes through the code and replaces
|
||
|
|
;;; #:BQ-LIST with LIST, #:BQ-APPEND with APPEND, and so on.
|
||
|
|
;;; #:BQ-CLOBBERABLE is simply eliminated (a call to it being
|
||
|
|
;;; replaced by its argument). #:BQ-LIST* is replaced by either
|
||
|
|
;;; LIST* or CONS (the latter is used in the two-argument case,
|
||
|
|
;;; purely to make the resulting code a tad more readable).
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-completely-process (x)
|
||
|
|
(let ((raw-result (bq-process x)))
|
||
|
|
(bq-remove-tokens (if *bq-simplify*
|
||
|
|
(bq-simplify raw-result)
|
||
|
|
raw-result))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-process (x)
|
||
|
|
(cond ((atom x)
|
||
|
|
(list *bq-quote* x))
|
||
|
|
((eq (car x) 'backquote)
|
||
|
|
(bq-process (bq-completely-process (cadr x))))
|
||
|
|
((eq (car x) *comma*) (cadr x))
|
||
|
|
((eq (car x) *comma-atsign*)
|
||
|
|
(error ",@~S after `" (cadr x)))
|
||
|
|
((eq (car x) *comma-dot*)
|
||
|
|
(error ",.~S after `" (cadr x)))
|
||
|
|
(t (do ((p x (cdr p))
|
||
|
|
(q '() (cons (bracket (car p)) q)))
|
||
|
|
((atom p)
|
||
|
|
(cons *bq-append*
|
||
|
|
(nreconc q (list (list *bq-quote* p)))))
|
||
|
|
(when (eq (car p) *comma*)
|
||
|
|
(unless (null (cddr p)) (error "Malformed ,~S" p))
|
||
|
|
(return (cons *bq-append*
|
||
|
|
(nreconc q (list (cadr p))))))
|
||
|
|
(when (eq (car p) *comma-atsign*)
|
||
|
|
(error "Dotted ,@~S" p))
|
||
|
|
(when (eq (car p) *comma-dot*)
|
||
|
|
(error "Dotted ,.~S" p))))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; This implements the bracket operator of the formal rules.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bracket (x)
|
||
|
|
(cond ((atom x)
|
||
|
|
(list *bq-list* (bq-process x)))
|
||
|
|
((eq (car x) *comma*)
|
||
|
|
(list *bq-list* (cadr x)))
|
||
|
|
((eq (car x) *comma-atsign*)
|
||
|
|
(cadr x))
|
||
|
|
((eq (car x) *comma-dot*)
|
||
|
|
(list *bq-clobberable* (cadr x)))
|
||
|
|
(t (list *bq-list* (bq-process x)))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; This auxiliary function is like MAPCAR but has two extra
|
||
|
|
;;; purposes: (1) it handles dotted lists; (2) it tries to make
|
||
|
|
;;; the result share with the argument x as much as possible.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun maptree (fn x)
|
||
|
|
(if (atom x)
|
||
|
|
(funcall fn x)
|
||
|
|
(let ((a (funcall fn (car x)))
|
||
|
|
(d (maptree fn (cdr x))))
|
||
|
|
(if (and (eql a (car x)) (eql d (cdr x)))
|
||
|
|
x
|
||
|
|
(cons a d)))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; This predicate is true of a form that when read looked
|
||
|
|
;;; like %@foo or %.foo.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-splicing-frob (x)
|
||
|
|
(and (consp x)
|
||
|
|
(or (eq (car x) *comma-atsign*)
|
||
|
|
(eq (car x) *comma-dot*))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
|
||
|
|
;;; This predicate is true of a form that when read
|
||
|
|
;;; looked like %@foo or %.foo or just plain %foo.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-frob (x)
|
||
|
|
(and (consp x)
|
||
|
|
(or (eq (car x) *comma*)
|
||
|
|
(eq (car x) *comma-atsign*)
|
||
|
|
(eq (car x) *comma-dot*))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; The simplifier essentially looks for calls to #:BQ-APPEND and
|
||
|
|
;;; tries to simplify them. The arguments to #:BQ-APPEND are
|
||
|
|
;;; processed from right to left, building up a replacement form.
|
||
|
|
;;; At each step a number of special cases are handled that,
|
||
|
|
;;; loosely speaking, look like this:
|
||
|
|
;;;
|
||
|
|
;;; (APPEND (LIST a b c) foo) => (LIST* a b c foo)
|
||
|
|
;;; provided a, b, c are not splicing frobs
|
||
|
|
;;; (APPEND (LIST* a b c) foo) => (LIST* a b (APPEND c foo))
|
||
|
|
;;; provided a, b, c are not splicing frobs
|
||
|
|
;;; (APPEND (QUOTE (x)) foo) => (LIST* (QUOTE x) foo)
|
||
|
|
;;; (APPEND (CLOBBERABLE x) foo) => (NCONC x foo)
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-simplify (x)
|
||
|
|
(if (atom x)
|
||
|
|
x
|
||
|
|
(let ((x (if (eq (car x) *bq-quote*)
|
||
|
|
x
|
||
|
|
(maptree #'bq-simplify x))))
|
||
|
|
(if (not (eq (car x) *bq-append*))
|
||
|
|
x
|
||
|
|
(bq-simplify-args x)))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-simplify-args (x)
|
||
|
|
(do ((args (reverse (cdr x)) (cdr args))
|
||
|
|
(result
|
||
|
|
nil
|
||
|
|
(cond ((atom (car args))
|
||
|
|
(bq-attach-append *bq-append* (car args) result))
|
||
|
|
((and (eq (caar args) *bq-list*)
|
||
|
|
(notany #'bq-splicing-frob (cdar args)))
|
||
|
|
(bq-attach-conses (cdar args) result))
|
||
|
|
((and (eq (caar args) *bq-list**)
|
||
|
|
(notany #'bq-splicing-frob (cdar args)))
|
||
|
|
(bq-attach-conses
|
||
|
|
(reverse (cdr (reverse (cdar args))))
|
||
|
|
(bq-attach-append *bq-append*
|
||
|
|
(car (last (car args)))
|
||
|
|
result)))
|
||
|
|
((and (eq (caar args) *bq-quote*)
|
||
|
|
(consp (cadar args))
|
||
|
|
(not (bq-frob (cadar args)))
|
||
|
|
(null (cddar args)))
|
||
|
|
(bq-attach-conses (list (list *bq-quote*
|
||
|
|
(caadar args)))
|
||
|
|
result))
|
||
|
|
((eq (caar args) *bq-clobberable*)
|
||
|
|
(bq-attach-append *bq-nconc* (cadar args) result))
|
||
|
|
(t (bq-attach-append *bq-append*
|
||
|
|
(car args)
|
||
|
|
result)))))
|
||
|
|
((null args) result)))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun null-or-quoted (x)
|
||
|
|
(or (null x) (and (consp x) (eq (car x) *bq-quote*))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; When BQ-ATTACH-APPEND is called, the OP should be #:BQ-APPEND
|
||
|
|
;;; or #:BQ-NCONC. This produces a form (op item result) but
|
||
|
|
;;; some simplifications are done on the fly:
|
||
|
|
;;;
|
||
|
|
;;; (op '(a b c) '(d e f g)) => '(a b c d e f g)
|
||
|
|
;;; (op item 'nil) => item, provided item is not a splicable frob
|
||
|
|
;;; (op item 'nil) => (op item), if item is a splicable frob
|
||
|
|
;;; (op item (op a b c)) => (op item a b c)
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-attach-append (op item result)
|
||
|
|
(cond ((and (null-or-quoted item) (null-or-quoted result))
|
||
|
|
(list *bq-quote* (append (cadr item) (cadr result))))
|
||
|
|
((or (null result) (equal result *bq-quote-nil*))
|
||
|
|
(if (bq-splicing-frob item) (list op item) item))
|
||
|
|
((and (consp result) (eq (car result) op))
|
||
|
|
(list* (car result) item (cdr result)))
|
||
|
|
(t (list op item result))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; The effect of BQ-ATTACH-CONSES is to produce a form as if by
|
||
|
|
;;; `(LIST* ,@items ,result) but some simplifications are done
|
||
|
|
;;; on the fly.
|
||
|
|
;;;
|
||
|
|
;;; (LIST* 'a 'b 'c 'd) => '(a b c . d)
|
||
|
|
;;; (LIST* a b c 'nil) => (LIST a b c)
|
||
|
|
;;; (LIST* a b c (LIST* d e f g)) => (LIST* a b c d e f g)
|
||
|
|
;;; (LIST* a b c (LIST d e f g)) => (LIST a b c d e f g)
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-attach-conses (items result)
|
||
|
|
(cond ((and (every #'null-or-quoted items)
|
||
|
|
(null-or-quoted result))
|
||
|
|
(list *bq-quote*
|
||
|
|
(append (mapcar #'cadr items) (cadr result))))
|
||
|
|
((or (null result) (equal result *bq-quote-nil*))
|
||
|
|
(cons *bq-list* items))
|
||
|
|
((and (consp result)
|
||
|
|
(or (eq (car result) *bq-list*)
|
||
|
|
(eq (car result) *bq-list**)))
|
||
|
|
(cons (car result) (append items (cdr result))))
|
||
|
|
(t (cons *bq-list** (append items (list result))))))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
;;; Removes funny tokens and changes (#:BQ-LIST* a b) into
|
||
|
|
;;; (CONS a b) instead of (LIST* a b), purely for readability.
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
(defun bq-remove-tokens (x)
|
||
|
|
(cond ((eq x *bq-list*) 'list)
|
||
|
|
((eq x *bq-append*) 'append)
|
||
|
|
((eq x *bq-nconc*) 'nconc)
|
||
|
|
((eq x *bq-list**) 'list*)
|
||
|
|
((eq x *bq-quote*) 'quote)
|
||
|
|
((atom x) x)
|
||
|
|
((eq (car x) *bq-clobberable*)
|
||
|
|
(bq-remove-tokens (cadr x)))
|
||
|
|
((and (eq (car x) *bq-list**)
|
||
|
|
(consp (cddr x))
|
||
|
|
(null (cdddr x)))
|
||
|
|
(cons 'cons (maptree #'bq-remove-tokens (cdr x))))
|
||
|
|
(t (maptree #'bq-remove-tokens x))))
|
||
|
|
</pre><P>
|
||
|
|
<P>
|
||
|
|
Suppose that we first make the following definitions:
|
||
|
|
<P>
|
||
|
|
<P><pre>
|
||
|
|
(setq q '(r s))
|
||
|
|
(defun r (x) (reduce #'* x))
|
||
|
|
(setq r '(3 5))
|
||
|
|
(setq s '(4 6))
|
||
|
|
</pre><P>
|
||
|
|
<P>
|
||
|
|
Without simplification, the notation
|
||
|
|
<tt>$$(%%q)</tt> (which stands for <tt>``(,,q)</tt>)
|
||
|
|
is read as the expression
|
||
|
|
<P><pre>
|
||
|
|
(APPEND (LIST 'APPEND) (LIST (APPEND (LIST 'LIST) (LIST Q))))
|
||
|
|
</pre><P>
|
||
|
|
The value of this expression is
|
||
|
|
<P><pre>
|
||
|
|
(APPEND (LIST (R S)))
|
||
|
|
</pre><P>
|
||
|
|
and the value of this value is <tt>(24)</tt>. We conclude
|
||
|
|
that the net effect
|
||
|
|
of twice-evaluating <tt>``(,,q)</tt> is to take
|
||
|
|
the value <tt>24</tt> of the value <tt>(r s)</tt> of <tt>q</tt>
|
||
|
|
and plug it into the template <tt>( )</tt> to produce <tt>(24)</tt>.
|
||
|
|
<P>
|
||
|
|
With simplification, the notation
|
||
|
|
<tt>$$(%%q)</tt>
|
||
|
|
is read as the expression
|
||
|
|
<P><pre>
|
||
|
|
(LIST 'LIST Q)
|
||
|
|
</pre><P>
|
||
|
|
The value of this expression is
|
||
|
|
<P><pre>
|
||
|
|
(LIST (R S))
|
||
|
|
</pre><P>
|
||
|
|
and the value of this value is <tt>(24)</tt>.
|
||
|
|
Thus the two ways of reading <tt>$$(%%q)</tt> do not produce the
|
||
|
|
same expression-this we expected-but the values of the two ways are
|
||
|
|
different as well. Only the values of the values are the same.
|
||
|
|
In general, Common Lisp guarantees the result of
|
||
|
|
an expression with backquotes nested to depth <i>k</i> only after
|
||
|
|
<i>k</i> successive evaluations have been performed; the results after
|
||
|
|
fewer than <i>k</i> evaluations are implementation-dependent.
|
||
|
|
<P>
|
||
|
|
(Note that in the expression <tt>`(foo ,(process `(bar ,x)))</tt>
|
||
|
|
the backquotes are <i>not</i> doubly nested. The inner backquoted
|
||
|
|
expression occurs within the textual scope of a comma belonging
|
||
|
|
to the outer backquote. The correct way to determine the backquote
|
||
|
|
nesting level of any subexpression is to start a count at zero and
|
||
|
|
proceed up the S-expression tree, adding one for each backquote
|
||
|
|
and subtracting one for each comma. This is similar to the rule
|
||
|
|
for determining nesting level with respect to parentheses by scanning
|
||
|
|
a character string linearly, adding or subtracting one as parentheses
|
||
|
|
are passed.)
|
||
|
|
<P>
|
||
|
|
It is convenient to extend the ``=='' notation to handle multiple evaluation:
|
||
|
|
<i>x</i> == == <i>y</i> means that the expressions <i>x</i> and <i>y</i> may have
|
||
|
|
different results but they have the same results when twice evaluated.
|
||
|
|
Similarly, <i>x</i> == ==== <i>y</i> means that the values of the values of the
|
||
|
|
values of <i>x</i> and <i>y</i> are the same, and so on.
|
||
|
|
<P>
|
||
|
|
We can illustrate the differences between non-splicing and splicing
|
||
|
|
backquote inclusions quite concisely:
|
||
|
|
<P><pre>
|
||
|
|
$$(%%q) ==
|
||
|
|
(APPEND (LIST 'APPEND) (LIST (APPEND (LIST 'LIST) (LIST Q))))
|
||
|
|
== == (LIST 'LIST Q) => (LIST (R S)) => (24)
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
$$(%@%q) ==
|
||
|
|
(APPEND (LIST 'APPEND) (LIST Q))
|
||
|
|
== == Q => (R S) => 24
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
$$(%%@q) ==
|
||
|
|
(APPEND (LIST 'APPEND) (LIST (APPEND (LIST 'LIST) Q)))
|
||
|
|
== == (CONS 'LIST Q) => (LIST R S) => ((3 5) (4 6))
|
||
|
|
</pre><P>
|
||
|
|
<P><pre>
|
||
|
|
$$(%@%@q) ==
|
||
|
|
(APPEND (LIST 'APPEND) Q)
|
||
|
|
== == (CONS 'APPEND Q) => (APPEND R S) => (3 5 4 6)
|
||
|
|
</pre><P>
|
||
|
|
In each case I have shown both the unsimplified and simplified forms
|
||
|
|
and then traced the intermediate evaluations of the simplified form.
|
||
|
|
(Actually, the unsimplified forms do contain one simplification
|
||
|
|
without which they would be unreadable:
|
||
|
|
the <tt>nil</tt> that terminates each list has been systematically suppressed,
|
||
|
|
so that one sees <tt>(append <i>x</i> <i>y</i>)</tt> rather than
|
||
|
|
<tt>(append <i>x</i> <i>y</i> 'nil)</tt>.)
|
||
|
|
<P>
|
||
|
|
The following driver function is useful for tracing the behavior
|
||
|
|
of nested backquote syntax through multiple evaluations.
|
||
|
|
The argument <tt>ls</tt> is a list of strings; each string
|
||
|
|
will be processed by the reader (<tt>read-from-string</tt>).
|
||
|
|
The argument <tt>n</tt> is the number of evaluations desired.
|
||
|
|
<P><pre>
|
||
|
|
(defun try (ls &optional (n 0))
|
||
|
|
(dolist (x ls)
|
||
|
|
(format t "~&~A"
|
||
|
|
(substitute #\` #\$ (substitute #\, #\% x)))
|
||
|
|
(do ((form (macroexpand (read-from-string x)) (eval form))
|
||
|
|
(str " = " "~% => ")
|
||
|
|
(j 0 (+ j 1)))
|
||
|
|
((>= j n)
|
||
|
|
(format t str)
|
||
|
|
(write form :pretty t))
|
||
|
|
(format t str)
|
||
|
|
(write form :pretty t)))
|
||
|
|
(format t "~\&"))
|
||
|
|
</pre><P>
|
||
|
|
This driver routine makes it easdy to explore a large number of cases
|
||
|
|
systematically. Here is a list of examples that illustrate not only
|
||
|
|
the differences between <tt>,</tt> and <tt>,@</tt> but also their
|
||
|
|
interaction with <tt>'</tt>.
|
||
|
|
<P><pre>
|
||
|
|
(setq fools2 '(
|
||
|
|
"$$(foo %%p)"
|
||
|
|
"$$(foo %%@q)"
|
||
|
|
"$$(foo %'%r)"
|
||
|
|
"$$(foo %'%@s)"
|
||
|
|
"$$(foo %@%p)"
|
||
|
|
"$$(foo %@%@q)"
|
||
|
|
"$$(foo %@'%r)"
|
||
|
|
"$$(foo %@'%@s)"
|
||
|
|
))
|
||
|
|
</pre><P>
|
||
|
|
<P>
|
||
|
|
Consider this set of sample values:
|
||
|
|
<P><pre>
|
||
|
|
(setq p '(union x y))
|
||
|
|
(setq q '((union x y) (list 'sqrt 9)))
|
||
|
|
(setq r '(union x y))
|
||
|
|
(setq s '((union x y)))
|
||
|
|
</pre><P>
|
||
|
|
<P>
|
||
|
|
Here is what happened when I executed <tt>(try fools2 2)</tt> with
|
||
|
|
a non-<tt>nil</tt> value for the variable <tt>*bq-simplify*</tt> (to see
|
||
|
|
simplified forms). I have interpolated some remarks.
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,,p) = (LIST 'LIST ''FOO P)
|
||
|
|
=> (LIST 'FOO (UNION X Y))
|
||
|
|
=> (FOO (A B C))
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,,p</tt> means ``the value of <tt>p</tt> is a form;
|
||
|
|
use the value of the value of <tt>p</tt>.''
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,,@q) = (LIST* 'LIST ''FOO Q)
|
||
|
|
=> (LIST 'FOO (UNION X Y) (LIST 'SQRT 9))
|
||
|
|
=> (FOO (A B C) (SQRT 9))
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,,@q</tt> means ``the value of <tt>q</tt> is a list of forms;
|
||
|
|
splice the list of values of the elements of the value of <tt>q</tt>.''
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,',r) = (LIST 'LIST ''FOO (LIST 'QUOTE R))
|
||
|
|
=> (LIST 'FOO '(UNION X Y))
|
||
|
|
=> (FOO (UNION X Y))
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,',r</tt> means ``the value of <tt>r</tt> may be any object;
|
||
|
|
use the value of <tt>r</tt>
|
||
|
|
that is available at the time of first evaluation,
|
||
|
|
that is, when the outer backquote is evaluated.''
|
||
|
|
(To use the value of <tt>r</tt> that is available at the time of second evaluation,
|
||
|
|
that is, when the inner backquote is evaluated,
|
||
|
|
just use <tt>,r</tt>.)
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,',@s) = (LIST 'LIST ''FOO (CONS 'QUOTE S))
|
||
|
|
=> (LIST 'FOO '(UNION X Y))
|
||
|
|
=> (FOO (UNION X Y))
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,',@s</tt> means ``the value of <tt>s</tt> must be a singleton list of any object;
|
||
|
|
use the element of the value of <tt>s</tt>
|
||
|
|
that is available at the time of first evaluation,
|
||
|
|
that is, when the outer backquote is evaluated.''
|
||
|
|
Note that <tt>s</tt> must be a singleton list because it will be spliced
|
||
|
|
into a form <tt>(quote )</tt>, and the <tt>quote</tt> special form requires exactly
|
||
|
|
one subform to appear; this is generally true of the sequence <tt>',@</tt>.
|
||
|
|
(To use the value of <tt>s</tt> that is available at the time of second evaluation,
|
||
|
|
that is, when the inner backquote is evaluated,
|
||
|
|
just use <tt>,@s</tt>,in which case the list <tt>s</tt> is not restricted to be singleton,
|
||
|
|
or <tt>,(car s)</tt>.)
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,@,p) = (LIST 'CONS ''FOO P)
|
||
|
|
=> (CONS 'FOO (UNION X Y))
|
||
|
|
=> (FOO A B C)
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,@,p</tt> means ``the value of <tt>p</tt> is a form;
|
||
|
|
splice in the value of the value of <tt>p</tt>.''
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,@,@q) = (LIST 'CONS ''FOO (CONS 'APPEND Q))
|
||
|
|
=> (CONS 'FOO (APPEND (UNION X Y) (LIST 'SQRT 9)))
|
||
|
|
=> (FOO A B C SQRT 9)
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,@,@q</tt> means ``the value of <tt>q</tt> is a list of forms;
|
||
|
|
splice each of the values of the elements of the value of <tt>q</tt>,
|
||
|
|
so that many splicings occur.''
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,@',r) = (LIST 'CONS ''FOO (LIST 'QUOTE R))
|
||
|
|
=> (CONS 'FOO '(UNION X Y))
|
||
|
|
=> (FOO UNION X Y)
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,@',r</tt> means ``the value of <tt>r</tt> must be a list;
|
||
|
|
splice in the value of <tt>r</tt>
|
||
|
|
that is available at the time of first evaluation,
|
||
|
|
that is, when the outer backquote is evaluated.''
|
||
|
|
(To splice the value of <tt>r</tt> that is available at the time of second evaluation,
|
||
|
|
that is, when the inner backquote is evaluated,
|
||
|
|
just use <tt>,@r</tt>.)
|
||
|
|
<P><pre>
|
||
|
|
``(foo ,@',@s) = (LIST 'CONS ''FOO (CONS 'QUOTE S))
|
||
|
|
=> (CONS 'FOO '(UNION X Y))
|
||
|
|
=> (FOO UNION X Y)
|
||
|
|
</pre><P>
|
||
|
|
So <tt>,@',@s</tt> means ``the value of <tt>s</tt> must be a singleton list whose
|
||
|
|
element is a list;
|
||
|
|
splice in the list that is the element of the value of <tt>s</tt>
|
||
|
|
that is available at the time of first evaluation,
|
||
|
|
that is, when the outer backquote is evaluated.''
|
||
|
|
(To splice the element of the value of <tt>s</tt> that is available at the time of second evaluation,
|
||
|
|
that is, when the inner backquote is evaluated,
|
||
|
|
just use <tt>,@(car s)</tt>.)
|
||
|
|
<P>
|
||
|
|
I leave it to the reader to explore the possibilities of triply nested backquotes.
|
||
|
|
<P><pre>
|
||
|
|
(setq fools3 '(
|
||
|
|
"$$$(foo %%%p)" "$$$(foo %%%@q)"
|
||
|
|
"$$$(foo %%'%r)" "$$$(foo %%'%@s)"
|
||
|
|
"$$$(foo %%@%p)" "$$$(foo %%@%@q)"
|
||
|
|
"$$$(foo %%@'%r)" "$$$(foo %%@'%@s)"
|
||
|
|
"$$$(foo %'%%p)" "$$$(foo %'%%@q)"
|
||
|
|
"$$$(foo %'%'%r)" "$$$(foo %'%'%@s)"
|
||
|
|
"$$$(foo %'%@%p)" "$$$(foo %'%@%@q)"
|
||
|
|
"$$$(foo %'%@'%r)" "$$$(foo %'%@'%@s)"
|
||
|
|
"$$$(foo %@%%p)" "$$$(foo %@%%@q)"
|
||
|
|
"$$$(foo %@%'%r)" "$$$(foo %@%'%@s)"
|
||
|
|
"$$$(foo %@%@%p)" "$$$(foo %@%@%@q)"
|
||
|
|
"$$$(foo %@%@'%r)" "$$$(foo %@%@'%@s)"
|
||
|
|
"$$$(foo %@'%%p)" "$$$(foo %@'%%@q)"
|
||
|
|
"$$$(foo %@'%'%r)" "$$$(foo %@'%'%@s)"
|
||
|
|
"$$$(foo %@'%@%p)" "$$$(foo %@'%@%@q)"
|
||
|
|
"$$$(foo %@'%@'%r)" "$$$(foo %@'%@'%@s)"
|
||
|
|
))
|
||
|
|
</pre><P>
|
||
|
|
It is a pleasant exercise to construct values for <tt>p</tt>, <tt>q</tt>, <tt>r</tt>, and <tt>s</tt>
|
||
|
|
that will allow execution of <tt>(try fools3 3)</tt> without error.
|
||
|
|
|
||
|
|
<br><img align=bottom alt="change_end" src="gif/change_end.gif">
|
||
|
|
|
||
|
|
<P>
|
||
|
|
<br> <HR><A NAME=tex2html6284 HREF="node368.html"><IMG ALIGN=BOTTOM ALT="next" SRC="icons/next_motif.gif"></A> <A NAME=tex2html6282 HREF="clm.html"><IMG ALIGN=BOTTOM ALT="up" SRC="icons/up_motif.gif"></A> <A NAME=tex2html6276 HREF="node366.html"><IMG ALIGN=BOTTOM ALT="previous" SRC="icons/previous_motif.gif"></A> <A NAME=tex2html6286 HREF="node1.html"><IMG ALIGN=BOTTOM ALT="contents" SRC="icons/contents_motif.gif"></A> <A NAME=tex2html6287 HREF="index.html"><IMG ALIGN=BOTTOM ALT="index" SRC="icons/index_motif.gif"></A> <BR>
|
||
|
|
<B> Next:</B> <A NAME=tex2html6285 HREF="node368.html">References</A>
|
||
|
|
<B>Up:</B> <A NAME=tex2html6283 HREF="clm.html">Common Lisp the Language</A>
|
||
|
|
<B> Previous:</B> <A NAME=tex2html6277 HREF="node366.html"> Discussion</A>
|
||
|
|
<HR> <P>
|
||
|
|
<HR>
|
||
|
|
<P><ADDRESS>
|
||
|
|
AI.Repository@cs.cmu.edu
|
||
|
|
</ADDRESS>
|
||
|
|
</BODY>
|