Ok, after looking into how clojure resolves class literal symbols, it turns out it is already smart enough to recognize symbol names that start with "["; the problem is it's a pain to type such symbols into code.
Your code can work this way: user=> (defmethod mm (resolve (symbol "[B")) [b] (println "bytes")) #<MultiFn clojure.lang.mult...@6ee3849c> user=> (mm (.getBytes "a")) bytes nil In the end though, it turns out to be just another way to call (Class/ forName "[B"). On Mar 24, 6:54 pm, ataggart <alex.tagg...@gmail.com> wrote: > The java version would look like: > Class c = byte[].class; > > But since the clojure way to obtain a class is simply to use its name > literal, e.g.: > user=> (type String) > java.lang.Class > the only way to get the class of an array would involve either a > function call or a change to the reader. And java doesn't make it > easy since all array-class stuff, such as Class.isArray(), is native > code. The best I've come up with is: > user=> > (defn #^Class array-class > "Returns the Class of an array of component type c" > [#^Class c] > (when c (class (java.lang.reflect.Array/newInstance c 0)))) > user=> (array-class String) > [Ljava.lang.String; > which appears faster than using Class/forName with a created string. > But since one *can* type a literal byte array class in java, I'd > imagine there's some way to implement this with bytecode magic, rather > than going through java.lang.reflect.Array. > > On Mar 24, 3:39 pm, Frank Siebenlist <frank.siebenl...@gmail.com> > wrote: > > > > > Right - not very different from the (class (byte-array 1)) that I came up > > with in the mean time... all not very clojuresque. > > > -FS. > > > On Mar 24, 2010, at 3:03 PM, ataggart wrote: > > > > For type-hinting #^"[B" works, but for obtaining and passing the class > > > to defmethod, the best I can come up with is (Class/forName "[B"). > > > > On Mar 24, 11:02 am, Frank Siebenlist <frank.siebenl...@gmail.com> > > > wrote: > > >> The following repl session shows my attempt to dispatch a multimethod on > > >> "type": > > > >> ... > > >> user> (defmulti mm type) > > >> #'user/mm > > >> user> (type "a") > > >> java.lang.String > > >> user> (defmethod mm java.lang.String [s] (println "string")) > > >> #<MultiFn clojure.lang.mult...@41e3a0ec> > > >> user> (mm "a") > > >> string > > >> nil > > >> user> (type (.getBytes "a")) > > >> [B > > >> user> (defmethod mm [B [b] (println "bytes")) > > >> ; Evaluation aborted. > > >> user> (def ba-type (type (.getBytes "a"))) > > >> #'user/ba-type > > >> user> (defmethod mm ba-type [b] (println "bytes")) > > >> #<MultiFn clojure.lang.mult...@41e3a0ec> > > >> user> (mm (.getBytes "a")) > > >> bytes > > >> nil > > >> user> > > >> ... > > > >> It works easily for the string, but for a native java byte array, type > > >> (or class) gives me back this "[B", which I'm unable to use as a > > >> dispatch value for the defmethod. > > >> I can, however, assign the value to a reference and us that to dispatch > > >> on successfully - but that feels like a hack. > > > >> Is there a way to express the byte array type in a different way than > > >> "[B" that would work? > > > >> Thanks, Frank. > > > > -- > > > 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 > > > > To unsubscribe from this group, send email to > > > clojure+unsubscribegooglegroups.com or reply to this email with the words > > > "REMOVE ME" as the subject. -- 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 To unsubscribe from this group, send email to clojure+unsubscribegooglegroups.com or reply to this email with the words "REMOVE ME" as the subject.
