me: Maybe :y or %y could be the zapper (or some better-suited abstraction over setforms).

evanrmurphy: Maybe something with `=` in it?

Okay, =y it is. ^_^ I've implemented a working example at https://gist.github.com/772283 (and posted this thread as a link to it).

What place.arc provides is a reader macro such that =(a b c) reads as (place (a b c)) and results in a function that either gets or sets that place depending on whether an argument is passed to it.

Here are example implementations of '= and 'zap using this system:

  (mac = args
    `(do ,@(map (fn ((a b)) `(atomic (=,a ,b))) pair.args)))
  
  (def fn-zap (func place . rest)
    (do.place:apply func (do.place) rest))
  
  ; NOTE: The original evaluates the place before the function instead.
  (mac zap (func place . rest)
    `(atomic:fn-zap ,func =,place ,@rest))

The '= macro only provides an occasional benefit too. Most of the time, you can save a character by saying (=foo bar) instead of (= foo bar).

It seems this abstraction over 'setforms not only shortens a lot of arc.arc definitions but also removes the need for some of them altogether.

---

You should know that place.arc is a little hackish in at least two ways:

- First, it still allows the symbol '= to be interpreted in a non-reader-macro way, but it does so using hard-coded special cases, and it doesn't count ssyntax like '=.variable-to-set-to-nil. The hackishness only exists with '=. If another symbol beginning with #\= is in your code, such as '=mc (an example from Lathe), it'll break by design because it'll read as (place mc). Meanwhile, symbols like 'or= and 'expand=list are just fine; I didn't make #\= a symbol-terminating character.