2008-05-10
| 04:55 | esquesque | hello |
| 04:55 | esquesque | i was wondering if clojure had an equivalent of cl's return-from |
| 12:08 | Chouser | esquesque: I don't think so. There's been a related discussion recently: http://groups.google.com/group/clojure/browse_thread/thread/90eae0f570989254 |
| 12:14 | asbjxrn | Anyone got hints for how to write a macro that uses variable capture? |
| 12:20 | asbjxrn | I keep getting that "Can't use qualified name as parameter" error. |
| 12:21 | asbjxrn | http://paste.lisp.org/display/60547 for an example of what I want. |
| 15:28 | Chouser | asbjxrn: try ~(symbol "x") instead of just x |
| 15:46 | hoeck | asbjxrn: or ~'x |
| 15:56 | Chouser | no way! so much better. |
| 15:56 | Chouser | hoeck: thanks. :-) |
| 16:13 | hoeck | chouser: :) |
| 21:01 | asbjxrn | Chouser, I tried that. Same error, unfortunally. |
| 21:04 | rhickey | tried what? |
| 21:05 | asbjxrn | using ~'x instead of just x in my macro where I try to use variable capture. |
| 21:06 | asbjxrn | I posted on the group as well: http://groups.google.com/group/clojure/browse_thread/thread/b4102f39dcc28569# |
| 21:15 | rhickey | ~' is the way to expose symbols without resolution, so maybe something else is wrong? Can you paste it on paste.lisp.org? |
| 21:18 | asbjxrn | http://paste.lisp.org/display/60547#1 |
| 21:20 | asbjxrn | Hmm. The namespace isn't clean when I do this, but I hope there nothing that interferes |
| 21:22 | asbjxrn | Macroexpansion looks ok: http://paste.lisp.org/display/60547#2 |
| 21:24 | rhickey | env is your problem - you probably want a gensym, try env# |
| 21:25 | asbjxrn | Doh. |
| 21:37 | asbjxrn | Another question: How do I subclass a class? This feels like it should be elementary, but I haven figured it out. (Say I want to add a environment map to a JPanel, for example.) |
| 21:37 | rhickey | use proxy |
| 21:39 | rhickey | in order to get state, close over it |
| 21:39 | asbjxrn | I thought so, but I've only seen how to add/override func... Ah. another doh.... |
| 21:41 | rhickey | don't sweat it, it's a different way to look at things for sure |
| 21:42 | asbjxrn | Well, I'm coming from the lisp side of the fence, so closures shouldn't be a big conceptual step... |
| 21:43 | rhickey | right, but not the way you'd think of extending Java classes necessarily |
| 21:44 | rhickey | there is a newer way in SVN, and it does allow for a single final state field - see src/genclass.clj |