Many thanks Konrad, thats illuminating! Regards, alux
On 3 Jun., 13:36, Konrad Hinsen <konrad.hin...@fastmail.net> wrote: > On 3 Jun 2010, at 12:46, alux wrote: > > > Thats impressive. I'm still, hm, puzzled. Nice example of something > > that can only be done with syntax-quote, and outside of a macro. (Or > > can this be done without syntax-quote, and I just dont see it?) > > That particular example is easy to do without syntax-quote: > > (map (fn [s v] (list 'def s v)) '(a b c) '(1 2 3)) > > Syntax-quote shows its stregths for nested forms, which would be a > pain to construct step by step using list and cons - but it's always > possible. > > > Is there any predicate like (macro? my-macro), and, if we are at it > > (special-form? my-form) ? > > user> (defmacro macro? [x] `(get (meta (var ~x)) :macro false )) > #'user/macro? > user> (macro? and) > true > user> (macro? map) > false > > user> (special-symbol? 'def) > true > user> (special-symbol? 'map) > false > > Note that macro? must be defined as a macro because otherwise its > argument would be evaluated, which is not possible for a macro. > special-symbol? is built-in, it works by interrogating the compiler's > internal list of special forms. > > > We could even generalise map with the thing you just described!(?) > > It's not more general, it's different! > > > One more thing. I'm still heavily irritated about the expansion of > > '`(foo (bar ~a) ~b) > > > From clojure.org/reader, I learned "For Lists/Vectors/Sets/Maps, > > syntax-quote establishes a template of the corresponding data > > structure.", and from that, what I expected was > > (user/foo (user/bar a) b) > > > Why is this so utterly wrong, and where should I know? > > Your expression would first call user/bar with argument a, and then > feed that and b as arguments to a call to user/foo. But you don't want > to call the functions, you want to return a form that calls them when > passed to eval. Here is what you would come up with when doing the > equivalent of `(foo (bar ~a) ~b) by hand: > > (list 'user/foo (list 'user/bar a) b) > > What Clojure returns is equivalent, but more complicated because it > has to work with more complex input as well. > > Konrad. -- 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
