Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!sun-barr!cs.utexas.edu!uunet!mcvax!ukc!etive!aiai!jeff
From: jeff@aiai.ed.ac.uk (Jeff Dalton)
Newsgroups: comp.lang.lisp
Subject: Re: Use of backquote & macro question - a real use for &aux!
Message-ID: <407@skye.ed.ac.uk>
Date: 3 May 89 16:15:11 GMT
References: <2277@perseus.sw.mcc.com> <2292@perseus.sw.mcc.com>
Reply-To: jeff@aiai.UUCP (Jeff Dalton)
Organization: AIAI, University of Edinburgh, Scotland
Lines: 39

In article <2292@perseus.sw.mcc.com> rcp@perseus.sw.mcc.com (Rob Pettengill) writes:
>I want to thank everyone who sent in suggestions.  The solution I like best
>to my problem appeared in the response from donc@vaxa.isi.edu (Don Cohen).
>Useing &aux variables to handle the setup required to write the macro
>eliminates the need for the evaled let that bothered me in the original. 
>This is the first real use for &aux that I have seen!
>Here is my latest version:

Just to make sure there isn't some subtle confusion here, I think
I should point out that &AUX isn't the only straightforward way
to eliminate the evaluated LET.  In particular, the &AUX can be
replaced in a simple, mechanical way by a LET*.  A macro that
looks like this

(defmacro with-context
  ((state-holder state-binding-list) &body body
   &aux
   (state-vars ...)
   (state-vals ...)
   (state-temps ...))
  "This macro ..."
  `(let ... more backquoted stuff ...  ))

can be rewritten to look like this:

(defmacro with-context ((state-holder state-binding-list) &body body)
  "This macro ..."
  (let* ((state-vars ...)
         (state-vals ...)
         (state-temps ...))
    `(let ... more backquoted stuff ... )))

You need to use LET* rather than LET because the calculation of
STATE-TEMPS refers to STATE-VARS.  You could also use nested LETs.

This sort of thing can be done for any use of &AUX, so whether you
use &AUX or not is largely a matter of taste.

-- Jeff