#clojure – 2010-05-08

2010-05-08

00:55	cap11235	I'm trying to extend JPanel using gen-class, and I can't find a way to implement a constructor so I can add a KeyListener to it. From what I can tell, :init is only for setting up the :state reference for the class, and calling super's contructors. I tried having it have "this" as a parameter, but I get get an IllegalArgumentException. How can I add my KeyListener?
01:13	Drakeson	sometimes (maybe whenever a jar is actually downloaded) lein deps nags something like Exception in thread "main" java.lang.NoClassDefFoundError: clojure/lang/Agent
01:14	Drakeson	does that happen to you, too?
01:24	tomoj	never seen that before
01:24	tomoj	Drakeson: ^
01:28	Drakeson	tomoj: thanks
01:48	technomancy	Drakeson: that's a known bug; shouldn't affect functionality
01:48	technomancy	you can ctrl+c your way out of it
01:49	technomancy	perils of living on the edge
01:49	technomancy	Drakeson: it's not happening in a stable release, is it?
01:52	Drakeson	technomancy: no, but I cannot switch back to the stable. (I need :repositories, for instance).
01:53	Drakeson	technomancy: do you plan on integrating a `search' task?
01:53	technomancy	Drakeson: not in the next release, possibly in 1.3
01:55	Drakeson	how about a cache (as in apt-cache update), for instance in order to know which repository to find [the latest of] each artifact from
01:56	technomancy	Drakeson: it's a possibility; I haven't really thought it all the way through
01:56	technomancy	web-based search works well enough that it's not the highest priority for the next release for me
01:58	Drakeson	well, that is not just for search, it is to speedup the process of finding the repository that has the latest version of an artifact. If I am not mistaken, it currently hits all the repositories everytime lein deps is issued.
01:59	technomancy	yes, but changing that behaviour would involve hacking maven itself
01:59	Drakeson	aah, I see.
01:59	technomancy	there are a lot of ways you could be smarter than mvn's current behaviour obviously, but I don't think any of them are really doable without at least doubling lein's LOC count
02:00	technomancy	would love to be proven wrong though =)
02:00	Drakeson	Will seting up a local maven "proxy" help?
02:01	technomancy	yeah, definitely
02:01	technomancy	I mean, probably
02:02	technomancy	actually I wonder if the Agent classdef not found is a Clojure bug
02:03	replaca	late friday night in #clojure. hi, gang
02:03	technomancy	replaca: heyo
02:03	technomancy	replaca: as a pprint kind of guy you might be interested in this: http://github.com/technomancy/serializable-fn
02:04	replaca	technomancy: saw it. very cute!
02:04	technomancy	we hacked it up at last night's seattle clojure meeting
02:04	replaca	I've thought about doing that a bunch of times, but never actually sat down to work it all the way out
02:05	technomancy	it was way easier than I thought it'd be now that you have access to &form and &env
02:05	technomancy	I tried adapting clojure.core/fn to do it, but the weirdness of how that macro is defined stymied me
02:05	replaca	yeah, I was looking at that a week or so ago and there's a lot going on there
02:06	technomancy	the first hundred lines of core.clj have changed a lot since I last looked =)
02:06	technomancy	it was pretty illuminating to go over it though, shows pretty well how &form and &env are done
02:07	replaca	core.clj is kind of like the bible, it has to be studied and you also have to take context into account (since it unfolds in order)
02:07	replaca	I've been messing around with declare to keep it from squashing metadata
02:07	technomancy	that analogy works surprisingly well
02:09	replaca	in the beginning, Rich said "Let there be #'list and there was #'list"
02:10	technomancy	"and Rich divided the fns from the macros..."
02:11	technomancy	man, I think it's bad when I break things while hacking swank; I hate to think what it's like when you mix things up while redefining fn
02:12	replaca	yeah, but it breaks pretty quickly at least
02:15	replaca	it's hard to even have a conversation about git vs. perforce anymore. So much of what I do would be unimaginably annoying in perforce
04:33	fnuction	Does anyone know if ClojureCLR will work against .Net 4?
05:14	Borkdude	fnuction: no, I haven't got it working in my VS2008 whatsoever
05:14	Borkdude	I didn't mean no, it doesn't work on .NET 4, I meant: I don't know -- sorry ;)
05:16	fnuction	I had it working against 3.5 in VS 2010, then I broke it.
05:18	fnuction	Post on the Google group explains why..
05:32	nurv	Hi.
05:38	Borkdude	fnuction: I didn't follow the GG much, I just tried what was said on the website but I kept getting errors
05:39	Borkdude	Something about references it couldn't find, although Intellisense could find it
05:48	fnuction	Borkdude: There are lots of potential problems with finding multiple version of the assemblies etc.
05:49	fnuction	nurv: Hi.
05:50	Borkdude	yes, hello nurv
08:11	GeoffSK	can some suggest why http://pastebin.com/raw.php?i=hsqMPeru causes a java exception?
08:20	hoeck	GeoffSK: probably the #((echo name running))
08:20	hoeck	try #(echo name running) instead
08:21	hoeck	because the echo funktion returns nil and you dont want to call nil #(nil) <- throws NPE
08:24	GeoffSK	hoeck: thanks (again)
08:25	hoeck	GeoffSK: np :)
08:25	GeoffSK	btw, is this an acceptable way to tell a thread to exit?
09:27	tufflax	It's great that I can do (rest [1 2 3]) on vectors like that, but how can I use that as a vector again? The class of it is clojure.lang.PersistentVector$ChunkedSeq, so underneath somehow it is still a vector, so I should be able to "cast" it back or something, yes?
09:49	patrkris	tufflax: the answer may depend on what you're trying to do here
09:49	patrkris	usually, if you're just looping through a vector to do something to each item, then it doesn't matter that you're not getting a vector back
09:50	patrkris	but for cases where you need to get a subvector the aptly named subvec function may help you
10:07	patrkris	rhickey: if you are there, could you help me understand a few things about obstruction freedom?
10:12	zakwilson	I'm finding several tutorials and libraries related to Google App Engine with Clojure. Is there one library that's generally preferred over others?
10:19	SynrG	hmm. something's wrong with my vimclojure ... doesn't look like localleader is doing its stuff ...
10:20	SynrG	if i press \ then s, it just performs an 's'
10:20	SynrG	:p
10:25	patrkris	SynrG: have you started the nailgun server?
10:29	SynrG	yes
10:30	patrkris	and does the nailgun server have the necessary .jar-files in its classpath?
10:30	patrkris	you can check the classpath by invoking `ng ng-cp`
10:49	SynrG	patrkris: seems like a vim issue, not a nailgun issue
10:50	SynrG	pressing \ and any other key doesn't do anything. it's as if i hadn't pressed the key at all :p
10:52	patrkris	SynrG: I had the same symptoms. It seems that vimclojure does not map keys for starting the REPL and performing evaluation unless it detects that the nailgun server is started. Even though you may have set the localleader to something, I don't think it comes into effect unless some key combination uses it.
10:53	SynrG	aha
10:53	patrkris	SynrG: but of course, it could also be a Vim issue. Did you get syntax highlighting working? Just to get an indication whether you installed vimclojure's vim scripts correctly into ~/.vim
10:54	SynrG	syntax highlighting works fine
10:54	patrkris	Ok. And you unzipped the vim-stuff as per the instructions?
10:55	SynrG	yes.
10:56	patrkris	Ok. Then I'm pretty sure it has something to do with your nailgun setup. Btw, are you using a *nix or Windows?
10:56	SynrG	when i ask vim what the filetype is when i'm editing a file it says 'clojure'
10:56	SynrG	*nix
10:56	patrkris	ok, that's fine then
10:56	SynrG	debian, to be precise
10:56	patrkris	Ok. Do you have the nailgun client in your path?
10:56	SynrG	yes
10:57	patrkris	have you tried invoking ng ng-cp?
10:57	SynrG	yes. it gives reasonable results
10:57	SynrG	it includes clojure.jar, clojure-contrib.jar, vimclojure.jar
10:57	patrkris	Ok... hmmm...
10:57	SynrG	and jline.jar
10:57	patrkris	ok
10:58	patrkris	have you tried setting localleader to something else?
10:58	SynrG	yes
10:58	patrkris	sorry, it's called maplocalleader - but you probably know that
10:59	patrkris	and you have the g:clj_want_gorilla option set to 1, I guess?
10:59	SynrG	yes. i tried changing it after loading a .clj file and earlier, in .vimrc
10:59	SynrG	no. is that essential?
10:59	patrkris	yes :)
10:59	patrkris	as far as I remember, I don't think it is enabled by default
11:00	patrkris	try setting it in your vimrc and see if it works
11:00	patrkris	let g:clj_want_gorilla = 1
11:00	Borkdude	What does this stuff from the core function + actually do? {:inline (fn [x y] `(. clojure.lang.Numbers (add ~x ~y)))
11:00	Borkdude	:inline-arities #{2}}
11:00	Borkdude
11:01	SynrG	patrkris: still doesn't help
11:01	SynrG	things are suddenly much slower tho
11:01	patrkris	SynrG: strange
11:02	patrkris	you might want to check back here when kotarak is online
11:02	patrkris	or try the mailinglist
11:02	SynrG	k
11:03	patrkris	Borkdude: inlines calls to +...?
11:04	Borkdude	patrkris when does that happen? I mean, there is also a 'normal' definition for two parameters
11:04	patrkris	i.e. wherever you use +, a call to clojure.lang.Numbers/add is inserted instead during compilation
11:05	Borkdude	aha
11:05	patrkris	Borkdude: maybe the normal definition is used when you don't use + directly but instead refer to it through an indrection
11:05	patrkris	but i'm just guessing :)
11:06	SynrG	patrkris: hm. something changed. i stopped ng-server, exited bash, started a fresh session to ensure my environment was properly set up, edited my file again ...
11:07	Borkdude	patrkris: maybe you're right, when it is compiled the function with 2 args gets inlined, I just wondered what this inlining stuff is about
11:07	SynrG	patrkris: now i get a lot of error spew and syntax highlighting is not turned on
11:07	patrkris	SynrG: then try :e<CR>
11:07	patrkris	SynrG: hopefully you know that <CR> means the Enter key :)
11:08	patrkris	Borkdude: i think it's for the sake of performance
11:08	SynrG	Couldn't execute Nail! java.io.FileNotFoundException: Could not locate book_list__init.class or b
11:08	SynrG	ook_list.clj on classpath: at ...
11:08	SynrG	hey, where'd my classpath go? looks empty
11:08	patrkris	SynrG: try editing an empty .clj file
11:09	patrkris	e.g. :e myfile.clj
11:09	patrkris	and then try starting a REPL via <localleader>sr e.g. \sr
11:10	SynrG	patrkris: nope. same symptom as before
11:11	SynrG	:let maplocalleader
11:11	SynrG	E121: Undefined variable: maplocalleader
11:11	zakwilson	Using Compojure and ring.adapter.jetty, changes I make at the REPL are not reflected when I reload a page. I'd RTFM, but there doesn't really seem to be one.
11:11	patrkris	ok, try :let maplocalleader=";"
11:11	patrkris	SynrG: ^
11:12	SynrG	no effect
11:12	patrkris	SynrG: I like to use semicolon, but you can try any key you'd like
11:12	patrkris	ok
11:13	patrkris	SynrG: could you try and close vim, but keep the nailgun server running and then start vim on an emtpy clojure-file ... just to try it out
11:13	patrkris	SynrG: and remember to put a let maplocalleader=";" in your .vimrc
11:15	SynrG	patrkris: still no joy :/
11:15	patrkris	SynrG: damn
11:16	patrkris	SynrG: but it seems that the "gorilla" part of vimclojure has been activated in some sense
11:17	patrkris	SynrG: btw which version of vimclojure are you using - the bleeding edge version or the last release?
11:17	SynrG	,, which is the localleader for vimoutliner seems to work
11:17	clojurebot	java.lang.Exception: Unable to resolve symbol: which in this context
11:18	SynrG	shush, clojurebot
11:18	patrkris	:)
11:18	SynrG	sec ...
11:18	SynrG	2.1.2
11:18	patrkris	yeah, i think it's the one I'm using
11:19	patrkris	well ... I don't have any other suggestions right now
11:19	SynrG	yup. thanks for trying :)
11:29	SynrG	patrkris: hm. rebuilt and reinstalled the jars. seems to be cooperating now
11:29	SynrG	dunno what i did :/
11:29	SynrG	i still get error spew at the beginning
11:30	patrkris	SynrG: can you start a REPL now?
11:31	SynrG	apparently
11:31	SynrG	but ;ef gives:
11:31	SynrG	Couldn't execute Nail! java.lang.ClassNotFoundException: de.kotka.vimclojure.nails.Repl at java.n
11:31	SynrG	et.URLClassLoader$1.run(URLClassLoader.java:217) at java.security.AccessController.doPrivileged(N
11:31	SynrG	ative Method) at java.net.URLClassLoader.findClass(URLClassLoader.java:205)
11:31	SynrG	test.clj contains:
11:31	SynrG	(+ 2 2)
11:31	clojurebot	4
11:32	SynrG	thanks, clojurebot :p
11:32	SynrG	when did clojurebot stop requiring a prefix?
11:33	patrkris	dunno - I never actually used it :)
11:33	patrkris	do you know where to find instructions on using clojurebot? :)
11:33	patrkris	strange that upon invoking ;ef it says the Repl class is missing
11:34	patrkris	but are you able to enter and evaluate text inside the REPL?
11:35	SynrG	restarted ng-server again. restarted editing again
11:35	SynrG	looks fine now
11:35	SynrG	mystifying
11:36	SynrG	i try random things and things work ...
11:36	SynrG	i hate it when that happens :p
11:36	tufflax	don't we all
11:37	SynrG	so, my HOWTO:
11:37	SynrG	obtain wall
11:37	SynrG	beat head against wall
11:37	SynrG	repeat until unconscious, or things work
11:38	patrkris	:)
11:42	tufflax	patrkris: Thanks by the way, for answering my question earlier. I wasn't trying to do anything specific, I was just curious.
11:47	SynrG	patrkris: things are not working flawlessly, however
11:48	SynrG	it seems very temperamental
11:51	patrkris	SynrG: you could take a look at the leiningen nailgun plugin, if you use leiningen
11:51	patrkris	SynrG: it sets up the classpath for you
11:52	SynrG	hm. always restart ng-server each time you start vim again
11:52	SynrG	seems to be part of my problem
11:54	SynrG	the other is that sometimes my ; seems to be 'eaten' by whatever i did previously, and i don't know why
11:54	SynrG	ah, it's timing
11:55	SynrG	because these sequences are unfamiliar, if i wait too long while i try to think of what the sequence is ...
11:55	SynrG	it forgets i typed localleader
11:58	patrkris	SynrG: that's right - also had that problem :)
12:00	patrkris	SynrG: I hope you get it to work - see you later
12:08	Drakeson	how can I extend an existing interface?
12:10	Drakeson	is gen-interface the only option?
12:13	hoeck	Drakeson: deftype, proxy or genclass
12:14	hoeck	oh, sorry, that implements an existing iface
12:20	Borkdude	What's with rfirst? I see it on the Clojure cheatsheet, but it doesn't work in my repl
12:23	Drakeson	Borkdude: nfirst
12:25	Drakeson	there was a rest --> next transition. rfirst and rrest probable got renamed to nfirst and nnext
12:26	Drakeson	How does one put an "Integer []" or "I[" type hint on an argument?
12:26	Drakeson	#^ints apparetnly does not do it
12:28	tcrayford	have you tried #^Integer[] ?
12:28	Drakeson	yes
12:28	Drakeson	fails
12:29	DuneMan	or #^{:tag Integer[]}
12:29	DuneMan	?
12:31	DuneMan	wonders
12:31	Drakeson	,(defn foo [#^{:tag Integer[]} x] x) (foo 1)
12:31	clojurebot	DENIED
12:31	Drakeson	,(defn foo [#^{:tag Integer[]} x] x)
12:31	clojurebot	DENIED
12:31	Drakeson	lazy bot
12:32	Drakeson	DuneMan: anyway, doesn't work
12:36	DuneMan	http://groups.google.com/group/clojure/browse_thread/thread/9a276a1709a88eeb
12:39	Drakeson	thanks, "ints" works for some other places, but apparently non in definterface.
12:40	DuneMan	well, "ints" is int[], no?
12:45	DuneMan	cap11235: looks like :init to me
12:45	Drakeson	DuneMan: it seems it should be
12:46	DuneMan	Drakeson: So that wouldn't be Integer[]
12:46	DuneMan	so it shouldn't work.
12:47	Drakeson	I don't get it. why wouldn't (defn foo [#^ints xs] xs) work?
12:47	DuneMan	If you're giving it Integer[], int[] != Integer[]
12:48	DuneMan	but you could use #^objects foo, and then give a typehint to integer for all of the accesses.
12:48	Drakeson	I am not giving it anything yet. it fails even before that. (definterface foo-iface (#^ints bar [this]))
12:48	DuneMan	oh, weird.
12:48	DuneMan	I see
12:49	Drakeson	is definterface being depreciated?
12:52	DuneMan	(gen-interface :name "Foo" :methods ....)
12:52	DuneMan	dur
12:52	DuneMan	yeah
12:55	DuneMan	I mean... I'm pretty new to clojure... but definterface isn't in the docs... so maybe it was a 1.0 thing?
13:06	dakrone	what's the easiest way to convert (:a 1 :b 2 :c 3) into a map? I was using (apply hash-map ...), but it throws errors on duplicate keys
13:08	Drakeson	,(into {} (vec '(:a 1 :b 2 :c 3)))
13:08	clojurebot	java.lang.IllegalArgumentException: Don't know how to create ISeq from: clojure.lang.Keyword
13:08	Drakeson	,(into {} (map vec (partition 2 '(:a 1 :b 2 :c 3))))
13:08	clojurebot	{:a 1, :b 2, :c 3}
13:09	dakrone	Drakeson: that's it, thank you much
13:27	tufflax	If I want to save clojure objects/datastructures to files, should I use java's stuff? Do clojure datastrctures implement serializeable?
13:27	Borkdude	good question
13:28	rhudson	They implement Serializable in 1.2
13:30	tufflax	ok, thanks
13:31	zakwilson__	With compojure.html, (html (vector [:h3 "foo"] [:p "bar"] [:div "baz"])) is returning "<><p>bar</p><div>baz</div></>". Is this a bug, or user-error?
13:32	zakwilson__	(expected was "<h3>foo</h3><p>bar</p><div>baz</div>")
13:52	_brian2_	noob question> regarding reading in a serialized java object, in java you use type casting to the object to identify it, how does that work in clojure -> http://clojure.pastebin.com/ZUmyyxYx
13:55	candeller	hi - having a list of strings, i.e '("foo" "bar" ...) Id like to get a list of the characters '(\f \o \o \b \a \r ...) what would be the most idiomatic way?
13:56	rhudson	_brian2_: The object you read in already has the right class. Clojure is dynamically typed.
13:56	_brian2_	i get type cast exception
13:57	rhudson	What's the expression?
13:57	rhickey	,(apply concat ["abc" "def"])
13:57	clojurebot	(\a \b \c \d \e \f)
13:58	candeller	thanks, wouldnt have guessed that concat can return separate chars :o
13:58	rhickey	,(seq "abc")
13:58	clojurebot	(\a \b \c)
13:59	candeller	oh, why i havent thought about it, thanks
14:01	_brian2_	rhudson: just a sec
14:04	dakrone	how do you refer to protocol methods outside of the namespace they're declared in? http://gist.github.com/394685
14:05	_brian2_	rhudson: (def classifier2 ( .readObject objIn )) , (def compiledClassifier ( AbstractExternalizable/compile classifier2))
14:10	_brian2_	rhudson : the problem , I think, is that this object needs to be cast into another type to be accepted by the function I'm passing it ot
14:11	_brian2_	this object is a subclass of the object I need to cast it into
14:13	rhudson	Sorry _brian2_, was away. So it's com.aliasi.util.AbstractExternalizable ?
14:13	_brian2_	yes
14:14	_brian2_	lingpipe library
14:15	rhudson	And classifier2 implments Compilable?
14:16	_brian2_	just a sec
14:18	_brian2_	rhudson : no, however ..
14:20	arkahn	is the ants.clj program Rich Hickey wrote two or three years ago still a canonical example in the use of clojure concurrency or is there something newer/updated that's a better example?
14:20	_brian2_	rhudson : I'm putting the authors notes up
14:21	arkahn	... sometimes I feel clojure suffers from Perl's TMTOWTDI - I just want one good way ; )
14:22	arkahn	a la Python
14:23	_brian2_	rhudson : -> http://clojure.pastebin.com/nh1aDCgY
14:26	_brian2_	he starts with DynamicLMClassifier , trains it, writes it out as that type, then appears to cast it into LMClassifer , which does not, I think, implement Compilable
14:26	_brian2_	when he reads it in
14:29	rhudson	When the object gets serialized, the serialization includes the name (or some indicator) of the class of the object. When you read it back it, you're reconstituting the same class of object.
14:30	rhudson	In Java, a statically typed language, you have to say what the types of your variables are.
14:30	rhudson	So when you cast to LMClassifier, you're saying "through this variable I will only perform LMClassifier calls".
14:31	rhudson	But the object itself is unchanged.
14:31	_brian2_	But appears I cant cast it in clojure
14:35	rhudson	You shouldn't need to cast it in Clojure. There's only one "compile" method in AbstractExternalizable, which takes a Compilable, so the implementation should be looking at the object, seeing whether it implements Compilable, and making the call if so.
14:35	rhudson	So I don't understand why you're getting a class cast exception.
14:37	rhudson	You could try putting in a type hint -- (.../compile #^Compilable classifier2)
14:37	_brian2_	ok
14:39	rhudson	or, when you've read in the object, make sure you have the right one: e.g. (assert (instance? Compilable classifier2))
14:41	_brian2_	ok, this #^Compilable didnt do anything
14:41	rhudson	yeah, I didn't really expect it to
14:54	rhudson	_brian2_, another troubleshooting thing to try is printing out the class of the object after you've read it in: (println (class classifier2)).
14:54	rhudson	And you're using the same version of the library to write and read?
14:55	_brian2_	ok
14:55	_brian2_	yes
14:56	nurv	Hi.
14:57	_brian2_	(println (class classifier2)) , com.aliasi.classify.LMClassifier
14:59	rhudson	which is not a Compilable
14:59	_brian2_	hmm
15:00	_brian2_	ok, its my stupidity
15:01	_brian2_	thanks!
15:01	rhudson	you're welcome
15:06	Borkdude	What's the difference with # and #^ ?
15:06	Borkdude	no sorry ^ and #^
15:13	cschreiner	#^ is oldschool
15:22	Borkdude	Is there any more idiomatic way of getting a sequence from \a to \z? (map char (range (int \a) (inc (int \z))))
15:25	rhudson	,(seq "abcdefghijklmnopqrstuvwxyz")
15:25	clojurebot	(\a \b \c \d \e \f \g \h \i \j \k \l \m \n \o \p \q \r \s \t \u \v \w \x \y \z)
15:25	ataggart	lol
15:26	LauJensen	naah, idiomatic would be more like
15:26	LauJensen	$(-> (map char (take 25 (drop-while (partial < 121) (iterate dec 10e6)))) reverse)
15:26	rhudson	It's actually fewer characters...
15:26	sexpbot	result: (\a \b \c \d \e \f \g \h \i \j \k \l \m \n \o \p \q \r \s \t \u \v \w \x \y)
15:26	rhudson	Zebra's won't like that...
15:27	Borkdude	LauJensen: I thought reverse was not so idiomatic, but more for people who come from CL ;)
15:28	LauJensen	of ofc you're right, try making it work with apply interleave instead :)
15:32	Borkdude	yes, much better than using inc, counting down is much more spacy
15:33	nurv	$(map char (range 97 123))
15:33	sexpbot	result: clojure.lang.LazySeq@726b766e
15:35	LauJensen	$(doall (map char (range 97 123)))
15:35	sexpbot	result: clojure.lang.LazySeq@726b766e
15:35	ataggart	btw, there are more than 26 letters
15:35	LauJensen	hrmf
15:35	Borkdude	hmm, I turned on clojure-mode in my ERC buffer, that didn't work out ;)
15:36	Borkdude	$(take-while true (map char (range 97 123)))
15:36	sexpbot	java.lang.Boolean cannot be cast to clojure.lang.IFn
15:36	Borkdude	ah damn ;)
15:37	Borkdude	$(take-while (constantly true) (map char (range 97 123)))
15:37	sexpbot	result: clojure.lang.LazySeq@726b766e
15:37	Borkdude	nopr
15:38	Borkdude	What happens if you use seq on it?
15:39	Borkdude	$(seq (map char (range 97 123)))
15:39	sexpbot	result: (\a \b \c \d \e \f \g \h \i \j \k \l \m \n \o \p \q \r \s \t \u \v \w \x \y \z)
15:40	nurv	mmmm...
15:41	nurv	$```````,(+ 2 2)
15:41	sexpbot	Command not found. No entiendo lo que estás diciendo.
15:41	hiredman	that is just sexpbot's broken result printin
15:41	nurv	Somehow it's a stack overflow here.
15:41	hiredman	,(take-while (constantly true) (map char (range 97 123)))
15:41	clojurebot	(\a \b \c \d \e \f \g \h \i \j \k \l \m \n \o \p \q \r \s \t \u \v \w \x \y \z)
15:58	hiredman	,(squote '(A B C# (unsquote-splice [(squote A) 2 3]) C#))
15:58	clojurebot	(list (quote sandbox/A) (quote sandbox/B) (quote C__15055) (quote sandbox/A) 2 3 (quote C__15055))
16:02	rhudson	,#'squote
16:02	clojurebot	#'sandbox/squote
16:03	hiredman	I really need to break the (defmulti foo type) habit and start writting protocols
16:11	vIkSiT	hmm, I'm trying to write some code that takes in a sentence like - This 'is' a test _sentence_ (quoted word and underscored word), and then performs an operation on it.
16:12	vIkSiT	now imperatively, I'd have an array/string, run through each word in that sentence in a loop, do an "if quoted() or if underscored()" on the word, and then perform the relevant operation, and append the changes into the string
16:13	vIkSiT	but given immutable data structures - whats a good functional way to do it without resorting to atoms and derefs?
16:14	rhudson	Basically you want to produce a new sentence that's the appropriate transform of the old sentence
16:15	vIkSiT	exactly
16:16	rhudson	so roughly (for [w words] (if (changeit? w) (changeit w) w))
16:17	vIkSiT	hmm yes. I'd have two if conditions though
16:17	vIkSiT	or n, for the purposes of this question..
16:17	rhudson	then cond
16:17	vIkSiT	I assume that the (changeit w) returns the changed value, but i'd have to operate on this _changed_ value in the next form..
16:18	vIkSiT	let me paste what i have right now..
16:18	vIkSiT	whats a good clojure pastebin btw?
16:18	vIkSiT	lisppaste?
16:18	clojurebot	lisppaste8: url
16:19	rhudson	dunno; haven't done it myself yet
16:20	rhudson	Lotsa folks seem to use gist.github.org
16:20	vIkSiT	thats true
16:20	vIkSiT	http://paste.lisp.org/display/98969
16:21	vIkSiT	so I have 2 issues here : a) how do i specify a "default" value to cond, and b) how do i take the return value from (myfun w) and pass it to the next statement?
16:21	rhudson	(a) :else w
16:22	rhudson	(b) using 'for usually works out well
16:22	rhudson	(a) -- any "true" value serves, but :else is conventional
16:23	vIkSiT	rhudson, ah, how dyou mean (b) there?
16:24	hiredman	a,b,a ?
16:24	rhudson	something like (let [words (for [w words] (cond ....))] (use-rebound-words ...)
16:24	hiredman	shoudl be ba
16:24	rhudson	hiredman: re his "so I have two issues here"
16:25	hiredman	oh, I assumed you had a binary alphabet
16:25	vIkSiT	hehe
16:25	rhudson	01101010000101
16:25	hiredman	why use for in this case over map?
16:25	hiredman	abbababaaaabab
16:25	rhudson	I just like for better
16:26	vIkSiT	hmm, I was thinking about map as well (i use it for everything in python) - but I'm still not sure in either case how the return value of (myfun w) is going to be reused?
16:26	rhudson	'for reads better than 'map when the function is long
16:26	vIkSiT	lets say I have a vector ["a" "b" "c"]
16:27	vIkSiT	and a function, (defn myfn [x])
16:27	hiredman	just say F
16:27	vIkSiT	sure
16:27	rhudson	F
16:28	hiredman	you want to transform the elements of V with F
16:28	vIkSiT	now after (for [w words] (cond (F? x) (F x)
16:28	vIkSiT	(F x) returns the transformed value of x
16:28	hiredman	(transform <=> map)
16:28	vIkSiT	ah. so map will do an in-place transform?
16:29	Borkdude	,(doc update-in)
16:29	clojurebot	"([m [k & ks] f & args]); 'Updates' a value in a nested associative structure, where ks is a sequence of keys and f is a function that will take the old value and any supplied args and return the new value, and returns a new nested structure. If any levels do not exist, hash-maps will be created."
16:29	hiredman	vIkSiT: how could it?
16:29	hiredman	immutable
16:29	rhudson	No, the point is you get a new vector , and apply the next operation to that new vector
16:29	vIkSiT	yes exactly.
16:29	hiredman	no
16:29	hiredman	you get a new seq
16:30	vIkSiT	I got that. My confusion is - how do I pass that new seq to the next (F? x)
16:30	hiredman	just call the next fn on it
16:30	rhudson	one way is, say ( -> ["a" "b" "c"] transform1 transform2 )
16:30	Borkdude	this sounds more like a map thing
16:30	hiredman	,(apply str (map char (map inc (range 90 100))))
16:30	clojurebot	"[\\]^_`abcd"
16:31	vIkSiT	hmmm yes that does seem better
16:31	vIkSiT	let me try
16:31	hiredman	,(-> (range 90 100) (map inc) (map char) (apply str))
16:31	clojurebot	java.lang.ClassCastException: clojure.lang.LazySeq cannot be cast to clojure.lang.IFn
16:31	Borkdude	(map #(if this or that then ret this or that) your-seq)
16:31	hiredman	,(->> (range 90 100) (map inc) (map char) (apply str))
16:31	clojurebot	"[\\]^_`abcd"
16:31	Drakeson	is there a way to transform the name of a class when you import it?
16:31	Drakeson	(or rename)
16:32	hiredman	try using for with -> or ->>
16:32	hiredman	gets funky fast
16:33	hiredman	,(->> x (for [y (range 10) :let [x (* y y)]]))
16:33	clojurebot	(0 1 4 9 16 25 36 49 64 81)
16:33	vIkSiT	ah
16:33	vIkSiT	what IS ->>?
16:34	hiredman	->> is ->, but differnt
16:34	rhudson	,(doc ->>)
16:34	clojurebot	"([x form] [x form & more]); Threads the expr through the forms. Inserts x as the last item in the first form, making a list of it if it is not a list already. If there are more forms, inserts the first form as the last item in second form, etc."
16:34	rhudson	,(doc ->)
16:34	clojurebot	"([x] [x form] [x form & more]); Threads the expr through the forms. Inserts x as the second item in the first form, making a list of it if it is not a list already. If there are more forms, inserts the first form as the second item in second form, etc."
16:35	vIkSiT	ha
16:35	vIkSiT	never used either
16:35	rhudson	They are really useful
16:35	Plouj	,(+ 1 2 3)
16:35	clojurebot	6
16:36	Borkdude	it's just a rewrite of what you would otherwise nest
16:36	vIkSiT	right
16:37	rhudson	Anyway, my point was not so much syntactic, as that functionally what you do is produce a new sentence (not alter the old one), and feed that to the next operation
16:38	vIkSiT	sure
16:41	rhudson	The arrows are a convenient way to do that if the operations are short forms (like names), and has the advantage of reading the right way. (-> a f1 f2 f3) vs (f3 (f2 (f1 a)))
17:12	springify	Good evening. Has anybody here ever tried cuke4duke with lein-cuke and made it work?
17:20	LauJensen	Is there something in contrib which makes (mkdir "x/y/z.clj") create both x, y and z.clj provided they dont exist?
17:21	Drakeson	LauJensen: Don't know, but (.mkdirs (java.io.File. "x/y")) should create x/y/
17:22	LauJensen	Good enough, thanks
17:22	mmarczyk	Karmic -> Lucid: yet another painless upgrade :-)
17:23	nickpad	hi - wondering if anyone's come across this problem trying to run the labrepl project: http://paste.lisp.org/display/98972
17:23	LauJensen	mmarczyk: Go get that lottery ticket
17:23	mmarczyk	LauJensen: yup, on my way already :-)
17:24	mmarczyk	apparently there's a Clojure package in universe, but it's 1.0... got to find out who maintains it and maybe bump it to 1.1
17:29	Plouj	hi
17:30	Plouj	so, I still can't figure out why swank/slime only works right after I install it in emacs with elpa, but fails with "Lisp connection closed unexpectedly: connection broken by remote peer" on subsequent emacs restarts.
17:49	Plouj	actually, sometimes it works (connects to addr=0.0.0.0/0.0.0.0,port=0,localport=53077) once every 5 times I try to restart and do M-x slime
17:55	Borkdude	mfex: here?
17:58	Plouj	what would you guys recommend as a rest for learning clojure for someone who isn't new to LISP and programming in general
18:00	Borkdude	as a rest?
18:02	Plouj	erm
18:02	Plouj	I don't know what I meant to type :)
18:02	Plouj	I guess as a "good book"
18:03	Plouj	I'm looking for a book/tutorial to learn clojure
18:03	Plouj	I'm not sure if reading the clojure version of sicp is that helpful for practicing concurrent programming
18:03	rhudson	Stuart Halloway's book Programming Clojure is good
18:04	dnolen	Plouj: sounds like Joy of Clojure is going to be pretty good. Programming Clojure is good too, though a lot of 1.2 stuff is missing from it.
18:05	Plouj	is joy of clojure more up to date, dnolen?
18:06	dnolen	Plouj: seems like it
18:06	rhudson	Joy of Clojure looks good so far -- 7 chapters are out in early access.
18:08	vIkSiT	does anyone cover protocols (outside of the video?)
18:08	rhudson	Neither book has exercise material. projecteuler.net and rosettacode.org are good places to find some interesting problems.
18:08	vIkSiT	There's also codingkata.org
18:09	rhudson	Haven't seen that one; I'll take a look
18:09	vIkSiT	(PE is most mathematical et al; I've found ck has some interesting stuff to get used to a language)
18:09	vIkSiT	mostly*
18:09	Plouj	project euler is hard :)
18:10	vIkSiT	Plouj, depends :) some of the "hardness" comes from not being familiar with certain concepts for instance.
18:11	vIkSiT	,(doc ->)
18:11	clojurebot	"([x] [x form] [x form & more]); Threads the expr through the forms. Inserts x as the second item in the first form, making a list of it if it is not a list already. If there are more forms, inserts the first form as the second item in second form, etc."
18:11	vIkSiT	rhudson, what are the -> and ->> operators called?
18:12	rhudson	Not sure they have an official name; I think of them as the arrow operators
18:12	Plouj	vIkSiT: yeah, it requires more concentrated thinking than I can afford while learning a language at the same time
18:12	The-Kenny	vIkSiT: Some people call -> "thread".
18:13	The-Kenny	(I forgot the name for the ->>)
18:13	vIkSiT	Plouj, agreed. I'd say PE/SPOJ/TC et al are to finetune language skills rather than learn; codingkata et al help do the latter pretty well
18:13	vIkSiT	(spoj - sphere online judge, tc - topcoder)
18:13	vIkSiT	they're also more algorithmic in nature
18:13	vIkSiT	The-Kenny, ah thanks
18:14	The-Kenny	,(doc ->>)
18:14	clojurebot	"([x form] [x form & more]); Threads the expr through the forms. Inserts x as the last item in the first form, making a list of it if it is not a list already. If there are more forms, inserts the first form as the last item in second form, etc."
18:14	rhudson	Rosettacode is a mixed bag. One advantage is that you see the same problem solved in a lot of different languages
18:14	vIkSiT	I was trying to find some example uses of -> and ->>
18:14	vIkSiT	but obviously google is horrendous in finding some
18:14	technomancy	vIkSiT: "thrush" is one name that I think encompasses -> and ->>
18:14	The-Kenny	,(-> 1 (+ 5) (- 10))
18:14	clojurebot	-4
18:14	technomancy	someone mentioned "stitch", which I think is much better than "thread"
18:15	The-Kenny	(- (+ 1 5) 10)
18:15	ataggart	-> "nest" since that's what it does
18:15	technomancy	hiredman: have you thought about changing the code invocation prefix for clojurebot? I get confused since clojurebot treats , like clojure treats ~
18:15	vIkSiT	technomancy, ah thanks. ironically a google of "clojure thrush" brings me to #clojure logs from 09 with the _exact_ same question being answered :)
18:16	technomancy	vIkSiT: http://debasishg.blogspot.com/2010/04/thrush-in-clojure.html <= a nice explanation
18:16	sexpbot	"Ruminations of a Programmer: Thrush in Clojure"
18:16	vIkSiT	ah. so -> takes 1, adds it as the second item in (+ 5 1), then -> takes the result of that and puts it as the second eleemnt in (- 10)
18:17	vIkSiT	technomancy, ah thanks. let me check
18:17	The-Kenny	vIkSiT: yes, and second element means after the function
18:18	rhudson	,(-> 7 (- 3))
18:18	clojurebot	4
18:18	rhudson	,(->> 7 (- 3))
18:18	clojurebot	-4
18:18	vIkSiT	hmm, after the function, as in?
18:18	vIkSiT	ah. -> implies (- 7 3), and ->> (- 3 7)
18:18	rhudson	right
18:19	vIkSiT	gotcha
18:20	rhudson	-> is especially useful in Java interop, since the standard Clojure form for a method call is (.method object args)
18:20	ataggart	-> is useful for nested operations on a value. ->> id useful for nested operations on a collection of values.
18:25	rhudson	I recently wanted to get a line sequence from a URL, so first I got an InputStream (with-open [istream (-> drafts-url URL. .openStream)]
18:25	rhudson	then
18:25	rhudson	(let [lnseq (-> istream InputStreamReader. BufferedReader. line-seq)]
18:37	DuneMan	I find -> so hard to read.
18:40	DuneMan	probably would get used to it if I used it.
18:40	DuneMan	*switches some code to use it
18:45	ataggart	its a nice replacement for deeply nested stuff which can be harder to read, e.g., (foo (bar (baz x) y) z) (-> x baz (bar y) (foo z))
18:45	ataggart	but if you don't like it, don't use it
18:46	ataggart	tmtowtdi
19:03	vIkSiT	hmm is <- an operator too?
19:03	vIkSiT	http://nathanmarz.com/blog/cascalog-news-feed/ uses it all over the place;
19:03	sexpbot	"News Feed in 38 lines of code using Cascalog — thoughts from the red planet"
19:04	rhudson	,(doc <-)
19:04	clojurebot	"clojure.contrib.datalog.rules/<-;[[hd & body]]; Build a datalog rule. Like this: (<- (:head :x ?x :y ?y) (:body-1 :x ?x :y ?y) (:body-2 :z ?z) (not! :body-3 :x ?x) (if > ?y ?z))"
19:04	vIkSiT	weird. why didnt that work for me?
19:05	vIkSiT	aah. c.c.datalog
19:05	vIkSiT	damn. the doc function should allow queries for ALL modules
19:05	vIkSiT	not just those used or required in the current namespace
19:05	rhudson	Yeah, clojurebot already has all the contrib libs
19:06	vIkSiT	is there a way to force doc to do that btw?
19:06	vIkSiT	useful if no web connection/IRc open!
19:06	rhudson	not to my knowledge
19:06	vIkSiT	oh well :)
19:25	somnium	,(apply assoc {} :a :b [1])
19:25	clojurebot	{1 nil, :a :b}
19:25	somnium	Im guessing its doing [[x y] [1]] to get the nil, would it make sense to throw there? (just took me a little bit to track down)
19:32	somnium	hrm, adapting it to conj instead of assoc makes it throw
20:04	dnolen	,(assoc {} :a :b 1)
20:04	clojurebot	{1 nil, :a :b}
20:07	somnium	,(assoc {} 1)
20:07	clojurebot	java.lang.IllegalArgumentException: Wrong number of args passed to: core$assoc
20:08	somnium	my POLS has been violated :/
20:25	tomoj	wat
20:25	tomoj	assoc should error with an even number of args, no?
20:26	hiredman	just look at the implementation of assoc
20:26	hiredman	it should be least surprising that it works that way
20:26	tomoj	is it RT.assoc?
20:26	hiredman	I dunno
20:26	hiredman	I haven't look at it
20:26	hiredman	but I can imagine how it goes
20:26	tomoj	hmm
20:27	tomoj	I think I might see what you mean
20:27	hiredman	(defn assoc ([a b c] ...) ([a & keyvals] blahblah blah reduce or loop))
20:28	tomoj	just an arity thing then?
20:28	hiredman	(let [keyvals '(1)] (assoc {} (first keyvals) (second keyvals)))
20:29	hiredman	,(let [keyvals '(1)] (assoc {} (first keyvals) (second keyvals)))
20:29	clojurebot	{1 nil}
20:29	hiredman	it's an artifact of being a little sloppy with seq processing
20:30	somnium	so 'sloppy artifacts' should be unsurprising?
20:32	somnium	anyway, being aware of it its not a problem, just was the first time I recall encountering it
20:33	hiredman	no, but if you look at the function, and say "how would I implement this" and the way you would implement it could easily yield certain sloppy behaviour, why would you be surprised to see that behaviour?
20:35	somnium	hmm, ease of deduction = probability of existence :P
21:00	vIkSiT	hmm. bit stuck on doing this in idiomatic clojure - i have a string to which i want to make a transformation. "this _is_a good string". I'd like to replace everything within ** with the match of that word in another dictionary
21:00	vIkSiT	(there can be multiple words with ** - I can do single words with re-gsub, for instance)
21:00	vIkSiT	not sure how best to do it. Ideas?
21:07	somnium	vIkSiT: are all ** whitespace delimited?
21:07	hiredman	fnparse
21:07	vIkSiT	somnium, yes
21:08	vIkSiT	hiredman, seems too complex for a toy program really :)
21:08	vIkSiT	fnparse, that s..
21:10	somnium	vIkSiT: for easy you could tokenize it and then (map (fn [s] (if-let [s* (dict s)] s* s)) tokens)
21:10	vIkSiT	hmm, what is s* here?
21:11	somnium	s* is a non-nil result from (dict "string")
21:22	somnium	,(require '[clojure.contrib.string :as s])
21:22	clojurebot	java.io.FileNotFoundException: Could not locate clojure/contrib/string__init.class or clojure/contrib/string.clj on classpath:
21:25	somnium	,(require '[clojure.contrib.str-utils2 :as s])
21:25	clojurebot	nil
21:57	vIkSiT	btw, using a clojure string, how do i get it without the first and last characters?
21:57	vIkSiT	I can do nth but it only returns the first char itself
21:57	vIkSiT	nth i mean
21:59	cp2	,(drop-last (rest "foo bar baz"))
21:59	clojurebot	(\o \o \space \b \a \r \space \b \a)
21:59	cp2	er
21:59	cp2	,(apply str (drop-last (rest "foo bar baz")))
21:59	clojurebot	"oo bar ba"
22:00	vIkSiT	hmmm
22:00	vIkSiT	a bit complex, for something that could be done using a[0] and a[-1] in other languages?
22:00	cp2	sure, but there are many ways to achieve something
22:00	vIkSiT	true
22:00	cp2	and all it really boils down to is how you use it in your code
22:01	vIkSiT	well, what i'm trying to do is - got a list of tokens. if any of them have *s as first and last, i want to strip and call a function using it..
22:01	vIkSiT	i guess i could use regexps.
22:02	cp2	so, if a string is like foo, call foo?
22:02	vIkSiT	more like: if string is like foo, call (f foo)
22:02	cp2	oh yeah, i missed that the first time around
22:03	vIkSiT	,(second (re-find #"\$(\S+)\$" "$test$")
22:03	clojurebot	EOF while reading
22:03	vIkSiT	,(second (re-find #"\$(\S+)\$" "$test$"))
22:03	clojurebot	"test"
22:03	vIkSiT	or something like that.. (for $ or *s)
22:03	somnium	vIkSiT: you may prefer the java String methods .contains, .startsWith, .endsWith etc
22:03	vIkSiT	oh good point.
22:04	cp2	mhm
22:04	cp2	honestly, your solution will be clunky just because of the way you need to deal with these tokens
22:04	cp2	i dont think its that big of a deal though
22:05	vIkSiT	sigh yeah
22:17	vIkSiT	bbl
22:17	cp2	http://tinyurl.com/2dyt9z8
22:17	sexpbot	"Vintage Symbolics 3650 36xx LISP Machine As-Is - eBay (item 260597652691 end time Jun-04-10 09:34:15 PDT)"
22:18	hiredman	huh, very tempting
22:18	hiredman	but I think the jvm will have to do as a lisp machine for now
22:19	cp2	unfortunately :(
22:21	cp2	i would still love to get my hands on one of those symbolics keyboards, though
22:21	cp2	that is a lot more affordable, at least =P
22:37	cemerick	~max
22:37	clojurebot	maxine is http://research.sun.com/projects/maxine/
22:37	cemerick	wha?
22:37	cemerick	~max
22:37	clojurebot	maxine is http://research.sun.com/projects/maxine/
22:37	cemerick	guess he's not tracking channel population anymore
22:37	cemerick	clojurebot: what do you know?
22:37	clojurebot	Pardon?
22:37	cemerick	~brain
22:37	hiredman	clojurebot: max users?
22:37	clojurebot	max people is 283
22:37	clojurebot	brain dump is http://clj.thelastcitadel.com/clojurebot
22:38	hiredman	the brain dump is busted
22:38	cemerick	ah, bummer
22:38	cemerick	maxine is pretty interesting, tho :-)
22:38	hiredman	right
22:38	cemerick	There was a good interview with one of the researchers a while back.
22:39	cemerick	se radio, perhaps
23:11	_brian2_	noob question> regarding exception handling, I 'm using a clojure library to scrape web pages, but how do I handle error 400 http exceptions, (try ( fetch ... doesn't seem to work
23:43	sids	_brian2_: the HTTP 400 errors are not the same as Java exceptions (unless the lib does the mapping)
23:43	sids	_brian2_: which library are you using?
23:59	_brian2_	sids : I'm using enlive