On Tue, Feb 8, 2011 at 12:31 AM, CuppoJava <patrickli_2...@hotmail.com> wrote:
> Thanks for the reply Ken.
>
> "While the macro is executing "body" will be bound to an
> s-expression (basically, source code parsed to an AST but no further;
> source code as processed by the reader). "
>
> This is the part where it falls apart. "body" is actually bound to a
> single symbol.
>
> For example in the destructuring-let macro:
> (def a [1 2 3])
> (let ((x y z) a)
> (println x y z))
>
> "body" refers to the single symbol "a".
What?
No. The syntax for let is a vector, so it'd be something like
(def a [1 2 3])
(let [[x y z] a]
(println x y z))
and we would have:
binding-forms '[[x y z] a]
body '(println x y z)
Furthermore I don't see what you need the body for.
In the binding-forms, it's true that the destructured input is not
known yet. The macro must emit something like
(let* [x (first a)
xxx (rest a)
y (first xxx)
yyy (rest xxx)
z (first yyy)]
(println x y z))
in cases like this. In short, the code it emits into the let* bindings
does the destructuring.
This code could be made to do double duty as follows.
First, whenever it can emit an s-expression like '(first a) or '(rest
a) it can check a flag and decide to either emit the s-expression or
do the equivalent thing, e.g. (first a), immediately on its argument
a.
Second, the let macro can generate the let* binding vector by calling
the function with the flag set to emit-s-expressions.
Third, the user-usable function can generate input to (into {}
(partition 2 ...)) by calling the function with the flag set to do-it.
It boils down to a few cases, like replacing
`(first ~sym)
with (if emit-sexps? `(first ~sym) (first sym))
and similarly elsewhere in the code that gets abstracted out into the helper-fn.
Then let becomes
(defmacro let [bindings & body]
`(let* ~(helper-fn bindings true)
~@body))
or something similar and the user destructure function is
(defn destructure [bindoids]
(into {} (partition 2 (helper-fn bindoids false))))
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
