If you (set! *warn-on-reflection* true) you'll see that the first uses reflection while the second doesn't. This is (imo) a bug in the method resolution behavior, namely that the code path for looking up compiled calls is different from reflected calls.
I submitted a patch to fix both this divergence and to allow for more correct lookup behavior when it comes to primitive args: https://www.assembla.com/spaces/clojure/tickets/445 Note that with the patch applied both scenarios result in a "More than one matching method found" exception, which is (imo) the correct behavior. Feel free to bug Stu, et al., to get the patch applied. ;) On Oct 26, 7:44 am, Maks Romih <mak...@gmail.com> wrote: > Hi! > > I stumbled on a difference of behavior in using a variable vs. let > binding. It may even be a bug in Clojure. > > Using Java 1.6 and Clojure 1.1 or 1.2, if you execute the following > code > > <code> > (import > 'javax.xml.crypto.dsig.XMLSignatureFactory > 'javax.xml.crypto.dsig.Transform) > > (def fac (XMLSignatureFactory/getInstance "DOM")) > > (let [tf (. fac newTransform Transform/ENVELOPED nil)] > tf) > > (let [fac (XMLSignatureFactory/getInstance "DOM") > tf (. fac newTransform Transform/ENVELOPED nil)] > tf) > </code> > > the first let works OK while the second one breaks with stack trace > and the error "More than one matching method found: newTransform". > There are really two methods with this name and the second parameter > nil does not differentiate them enough, but I'd like to know why the > first call, with the fac as a variable, succeeds. Which method of the > two does it call? > > Maks. -- 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
