#clojure – 2015-10-31

2015-10-31

00:27	pyon	Let's say that (foo bar) returns a list. How do I instead return a lazy sequence, with the same contents as foo bar?
00:28	justin_smith	(lazy-seq (foo bar)) - but there is little advantage to this, the real advantage is to generating your result lazily
00:28	pyon	Well, in my case (foo bar) is actually a recursive call.
00:28	pyon	Wait, just in case - `map` itself returns a lazy sequence, right?
00:28	justin_smith	then instead of (cons h (foo arg)) do (lazy-seq (cons h (foo arg))) inside the body of foo
00:29	justin_smith	sure, map returns a lazy-seq
00:32	pyon	What I really want to delay is the recursive call itself.
00:32	pyon	Rather than a cons.
00:33	justin_smith	you need cons to construct the list
00:33	justin_smith	the lazy-seq will delay all evaluation inside it
00:35	pyon	In this situation, I'm almost sure that what needs to be delayed is the recursive call itself: https://www.refheap.com/111230
00:36	justin_smith	pyon: like I said, lazy-seq in that position delays both the cons and the recursive call
00:36	amalloy	it's easier to just wrap the whole body of the function in lazy-seq
00:36	amalloy	but you can indeed put it just about anywhere and it'd be fine
00:36	justin_smith	fair point
00:37	pyon	:-O
00:37	justin_smith	so yeah, like amalloy says, you can just do (defn mono [k xs] (lazy-seq (when-let ...)))
00:56	jhn	why does (merge-with concat {} {:a [1]}) => {:a [1]}
00:56	jhn	but (swap! (atom {}) merge-with concat {:a [1]}) =>
00:56	jhn	IllegalArgumentException contains? not supported on type: clojure.core$concat
00:59	amalloy	because that's (merge-with {} concat ...)
01:00	jhn	ah! that makes sense. ty. :)
01:03	jhn	is this the least ugly way to make this work?
01:03	jhn	(swap! (atom {}) #(merge-with concat % {:a [1]}))
01:05	pyon	Why is lazy-seq called lazy-seq, if it can be used on things other than sequences?
01:05	jhn	because seq is the name of the interface.
01:05	jhn	so you can use it with anything that is a seq
01:07	jhn	essentially, the "things other than sequences" you're referring to are actually sequences.
01:12	mungojelly	everything is everything and programming is just an extended meditation on how similar everything is to everything else
01:21	pyon	jhn: Ah!
01:21	pyon	Is there any way to delay computations that don't yield (the beginning of) a sequence?
01:22	rhg135	defer
01:22	pyon	Ah!
01:22	rhg135	Err, delay
01:23	rhg135	,(def v (delay (println clojureversion)))
01:23	clojurebot	#error {\n :cause "Unable to resolve symbol: clojureversion in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: clojureversion in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve...
01:24	rhg135	,(def v (delay (println clojure-version)))
01:24	clojurebot	#'sandbox/v
01:24	rhg135	,@v
01:24	clojurebot	{:major 1, :minor 8, :incremental 0, :qualifier alpha3}\n
03:28	pyon	Is there anything like a recursive macroexpand?
03:29	pyon	That is, macroexpand everything. Or, even better, only macroexpand specific macros that I tell it to expand.
03:45	TEttinger	pyon: yeah
03:46	TEttinger	http://clojure.github.io/clojure/clojure.walk-api.html#clojure.walk/macroexpand-all
03:47	TEttinger	,(clojure.walk/macroexpand-all (if (and true "yay") :hooray :boo))
03:47	TEttinger	,(clojure.walk/macroexpand-all (if (and true "yay") :hooray :boo)))
03:47	clojurebot	#error {\n :cause "clojure.walk"\n :via\n [{:type java.lang.ClassNotFoundException\n :message "clojure.walk"\n :at [java.net.URLClassLoader$1 run "URLClassLoader.java" 366]}]\n :trace\n [[java.net.URLClassLoader$1 run "URLClassLoader.java" 366]\n [java.net.URLClassLoader$1 run "URLClassLoader.java" 355]\n [java.security.AccessController doPrivileged "AccessController.java" -2]\n [java.net.U...
03:47	clojurebot	#error {\n :cause "clojure.walk"\n :via\n [{:type java.lang.ClassNotFoundException\n :message "clojure.walk"\n :at [java.net.URLClassLoader$1 run "URLClassLoader.java" 366]}]\n :trace\n [[java.net.URLClassLoader$1 run "URLClassLoader.java" 366]\n [java.net.URLClassLoader$1 run "URLClassLoader.java" 355]\n [java.security.AccessController doPrivileged "AccessController.java" -2]\n [java.net.U...
03:47	TEttinger	,(require '[clojure.walk])
03:47	clojurebot	nil
03:47	TEttinger	,(clojure.walk/macroexpand-all (if (and true "yay") :hooray :boo))
03:47	clojurebot	:hooray
03:47	TEttinger	,(clojure.walk/macroexpand-all '(if (and true "yay") :hooray :boo))
03:47	clojurebot	(if (let* [and__4224__auto__ true] (if and__4224__auto__ "yay" and__4224__auto__)) :hooray :boo)
03:49	pyon	TEttinger: Ah, nice, thanks. :-)
03:54	pyon	If I defined a function `foo' inside a macro or function `bar', then `foo' can only be used inside `bar', right?
03:54	pyon	define*
04:02	pyon	Is there anything like `when-let`, but accepting multiple bindings?
04:03	pyon	Also, what's the difference between `defn` and `letfn`?
04:03	pyon	Or, rather, why do we need them both?
04:15	hiredman	because clojure is not scheme
04:15	hiredman	def in clojure always creates a top level binding, defn expands to a def + fn
04:16	hiredman	letfn creates local bindings and the names are all in scope for the functions bound to the names
04:16	hiredman	(similar to letrec)
04:33	pyon	hiredman: ah!
05:24	zajira	i'm new to clojure, how I import http-kit into the lein REPL? I already added http-kit to my project.clj, but I get "No such namespace: http-kit"
05:47	macro-pyon	What's the convention for embedding code in docstrings?
05:47	macro-pyon	Markdown?
06:10	expez	macro-pyon: no real convention, but I've only ever seen people use MD.
06:11	expez	macro-pyon: It's more about what your tools can handle. CIDER will highlight stuff embedded with `` in docstrings and comments for example.
06:11	expez	embedded in*
06:12	expez	Is : a part of the keyword, or is it more correct to think of this as a token for the reader?
06:12	expez	I'm leaning toward the latter
06:12	expez	,(name :foo)
06:12	clojurebot	"foo"
06:13	expez	which I think should return ":foo", for the former interpretation to hold
06:14	expez	I'm asking because we're still highlighting the : with keyword face in clojure-mode, even though the package prefix is now in a different color. I'm starting to think this is a mistake.
06:14	vijaykiran	,[(keyword "a")(keyword :a)(keyword "a")]
06:14	clojurebot	[:a :a :a]
06:17	justin_smith	,(keyword ":do not do this")
06:17	clojurebot	::do not do this
06:30	WickedShell	Is the ^:const hint considered valid/used with fn's?
06:30	justin_smith	you can use it, bu afaik it won't do anything
06:31	WickedShell	Not surprised to hear that (or really bothered tbh)
07:09	pyon	Is there a better way to do this? https://www.refheap.com/111232
07:09	pyon	I've been wondering if I could use `loop` instead.
07:11	pyon	Errr, replace diff-loop with differentiate.
07:11	pyon	https://www.refheap.com/111234 fix'd
07:20	justin_smith	that seq call is redundant - the destructuring implies it
07:20	pyon	Ah!
07:21	justin_smith	otherwise, looks good to me (maybe you could avoid using the same lhs twice in the let binding).
07:22	justin_smith	,(let [[h & t] "hello"] [h t])
07:22	clojurebot	[\h (\e \l \l \o)]
07:23	pyon	Ah!
07:23	pyon	If let bindings are sequential, this should do what I want it to do.
07:23	pyon	But, yeah, I should probably rename the second `xss`.
07:24	justin_smith	yes they are sequential
07:29	pyon	Is there no way to make this code shorter? (fn [[i xs]] [i (cons x xs)]) seems too long for what it achieves.
07:31	justin_smith	#(update-in % 1 conj x) if xs is a list, and [i xs] is a vector
07:31	justin_smith	err
07:31	justin_smith	update, not update-in
07:31	pyon	Ah!
07:32	justin_smith	,(update [:a '(:b :c :d)] 1 conj :aa)
07:32	clojurebot	[:a (:aa :b :c :d)]
07:32	justin_smith	it breaks if they are not list and vector though
07:44	xeqi	pyon: I'd prolly write that as a variation of https://www.refheap.com/111236
07:44	pyon	Whoa, that's so much neater.
07:45	pyon	Alas, incorrect. :-\|
07:46	pyon	Also, constantly performing (take i) and (drop i) for all values of i... is a little tad inefficient, isn't it? :-O
07:46	xeqi	ack, pasted the wrong one. (drop (inc i) xs)
07:46	pyon	Ah, good.
07:50	xeqi	might be. But I could see it the same as conj-ing the list when the (map fun (map fun (map fun xss))) stack unwinds with a layer for each previous value in the original
07:50	xeqi	if I was performance sensitive, I'd assume xs is a vector and use (subvec ...)
07:51	pyon	What's subvec do?
07:51	pyon	Never mind. Me being stupid. :-\|
07:51	pyon	Can vectors be efficiently concatenated, though?
07:59	xeqi	`concat` returns a lazy seq that walks underneath, so depends on how you're going to use it later. An alternative is `into` (https://www.refheap.com/111237) which uses transients underneath to build a new sequence
08:00	xeqi	here's where I refer you to setup a suite with https://github.com/hugoduncan/criterium if this seems performance sensitive
08:00	pyon	Checking.
08:07	pyon	From [1 2 3 4 5], how do I get [[] [1] [1 2] [1 2 3] [1 2 3 4] [1 2 3 4 5]]? (not necessarily vectors)
08:07	pyon	And also [[1 2 3 4 5] [2 3 4 5] [3 4 5] [4 5] [5] []].
08:09	justin_smith	,(iterate pop [1 2 3 4 5])
08:09	clojurebot	([1 2 3 4 5] [1 2 3 4] [1 2 3] [1 2] [1] ...)
08:09	justin_smith	,(iterate rest [1 2 3 4 5])
08:09	clojurebot	([1 2 3 4 5] (2 3 4 5) (3 4 5) (4 5) (5) ...)
08:10	justin_smith	you'll want a test for when it ends up empty
08:10	justin_smith	or a take-while
08:10	pyon	Ah!
08:10	justin_smith	,(take-while seq (iterate rest [1 2 3 4 5]))
08:10	clojurebot	([1 2 3 4 5] (2 3 4 5) (3 4 5) (4 5) (5))
08:10	justin_smith	,(take-while seq (iterate pop [1 2 3 4 5]))
08:10	clojurebot	([1 2 3 4 5] [1 2 3 4] [1 2 3] [1 2] [1])
08:11	justin_smith	when it's not a vector, you can use but-last instead of pop, but pop is much better for vectors
08:11	pyon	Ah!
08:11	pyon	,(let [xs [1 2 3 4 5]] (map vector (range) xs (iterate pop xs) (iterate rest xs)))
08:11	clojurebot	([0 1 [1 2 3 4 5] [1 2 3 4 5]] [1 2 [1 2 3 4] (2 3 4 5)] [2 3 [1 2 3] (3 4 5)] [3 4 [1 2] (4 5)] [4 5 [1] (5)])
08:12	pyon	Wait, no, I need something like the reverse of (iterate pop xs).
08:13	justin_smith	well, we do have reverse
08:13	pyon	yeah, but that turns out to be O(n²).
08:14	justin_smith	why n squared instead of n?
08:15	pyon	I mean, for the thing. Each reverse is O(n).
08:15	pyon	But I need n reverses.
08:17	justin_smith	,(reduce conj [] [1 2 3 4 5])
08:17	clojurebot	[1 2 3 4 5]
08:17	justin_smith	err
08:17	justin_smith	,(reductions conj [] [1 2 3 4 5])
08:17	clojurebot	([] [1] [1 2] [1 2 3] [1 2 3 4] ...)
08:17	pyon	Why can I do (#(vector %1 %2 %1 %2) "foo" "bar"), but not (#([ %1 %2 %1 %2]) "foo" "bar") ?
08:17	pyon	Ah!
08:18	justin_smith	,'#([:a]) ; this should help
08:18	clojurebot	(fn* [] ([:a]))
08:19	pyon	Ah!
08:25	pyon	What's the function for getting the i-th element of a vector?
08:25	justin_smith	nth
08:25	justin_smith	with a vector you can also use get
08:25	pyon	Ah, sweet. :-)
08:26	justin_smith	,((juxt nth get) [:a :b :c] 1)
08:26	clojurebot	[:b :b]
08:26	justin_smith	,(get-in [:a [:b [:c [:d [:e]]]]] [1 1 1 0])
08:26	clojurebot	:d
08:26	justin_smith	,(get-in [:a [:b [:c [:d [:e]]]]] [1 1 1 1 0])
08:26	clojurebot	:e
08:30	pyon	I found a neat solution: https://www.refheap.com/111238 :-)
08:31	pyon	Well, actually, it's mostly based on what you guys suggested. :-)
08:31	TEttinger	,([0 1 2 3] 0)
08:31	clojurebot	0
08:31	TEttinger	,([0 1 2 3] 3)
08:31	clojurebot	3
08:31	TEttinger	,([0 1 2 30] 3)
08:31	clojurebot	30
08:31	TEttinger	vectors are functions of numerical indices too
08:31	pyon	:-O
08:32	pyon	That makes sense.
08:32	TEttinger	similar to maps and their keys
08:32	TEttinger	,({:a 1 :b 2} :b)
08:32	clojurebot	2
09:28	sobel	but keys are functions and indices are not, so you can write (:a {:a 1 :b 2}) => 1 but (1 [2 3 4]) yields an exception
09:29	sobel	just to note a little asymmetry there :)
09:49	pyon	What was the syntax for creating a record value? (not the record type)
09:50	pyon	Oh, there was a dot at the end of the record's name! I forgot that.
09:57	Glenjamin	there's also the clojure function ->Record
09:58	Glenjamin	,(do (defrecord Example [a b]) [(Example. 1 2) (->Example 1 2) (map->Example {:a 1 :b 2})])
09:58	clojurebot	[#sandbox.Example{:a 1, :b 2} #sandbox.Example{:a 1, :b 2} #sandbox.Example{:a 1, :b 2}]
10:10	pyon	Could someone please help me debug this macro? https://www.refheap.com/111241
10:10	pyon	When I run this in the REPL, I get the desired result: `[~@(mapcat (partial constructor-partial-derivative :r) node)]
10:10	pyon	But, for some reason, when I run it from within the macro, it doesn't work. :-\|
10:34	xeqi	pyon: in total-differential you want `map-indexed` instead of `keep-indexed` since you keep everything
10:34	pyon	Ah, thanks!
10:34	pyon	Also, I found the solution to the defderivative problem.
10:34	pyon	I needed an eval.
10:34	pyon	(eval base)
10:41	xeqi	pyon: can defderivative be a function? looks like it just evaluates everything and passes it through?
10:41	pyon	xeqi: defderivative isn't finished.
10:41	pyon	In the end, it will be a macro - a much more complex one than defpolynomial.
10:42	pyon	It's just that I needed to see something for debugging purposes.
11:34	pyon	Is there anything like map, but which leaves the first element untouched?
13:50	gfredericks	pyon: that'd be a pretty weird function to have built in I think, but it's easy to write yourself
13:51	gfredericks	,(defn map-rest [f coll] (if (empty? coll) coll (cons (first coll) (map f (rest coll)))))
13:51	clojurebot	#'sandbox/map-rest
13:51	gfredericks	,(map-rest inc (range 5))
13:51	clojurebot	(0 2 3 4 5)
13:51	gfredericks	,(map-rest dec ())
13:51	clojurebot	()
13:55	xeqi	,(map-rest dec nil)
13:55	clojurebot	nil
13:56	xeqi	ah, map returns () cause it uses lazy-seq
13:57	J_Arcane	,(first ())
13:57	clojurebot	nil
13:57	J_Arcane	,(rest ())
13:57	clojurebot	()
14:36	justin_smith	,(map (juxt rest next) [() [] nil])
14:36	clojurebot	([() nil] [() nil] [() nil])
14:53	xeqi	what function turns a vec of keys and a vec of vals into a map?
14:53	justin_smith	zipmap
14:53	justin_smith	,(zipmap [:a :b :c] [1 2 3])
14:53	clojurebot	{:a 1, :b 2, :c 3}
14:54	xeqi	(inc justin_smith)
14:54	xeqi	oh, right.. no lazybot
15:24	skoude	Noob question,, but I have a json parsed with chesire.. and it's like [ { employee: {firstname: "test", lastname: "test" }, employer: { companyid: "test"} } , {employee: {firstname: "test2", lastname: "test2}, employer: {companyid: "test2"} }].. So how can I get all of the firstnames?
15:25	skoude	I tried to use (get-in parsed_json["employee" "firstname"], but it does not work
15:27	justin_smith	if that's the real structure, it should be (get-in parsed_json [0 "employee" "firstname"])
15:28	justin_smith	unless you provided the keywordize option to cheshire, in that case change the strings to keywords
15:28	rhg135	or map it
15:28	justin_smith	oh, all the first names
15:28	justin_smith	missed that part, right you are rhg135
15:29	skoude	thanks, now I got it. I had the keywordize option so I use keywords and it works :)
15:36	skoude	hmm.. I used (cheshire.core/parse-string s true)) to parse the json.. and if I print the parsed json, it looks okay.. but still I get only nil when using: get-in parsed_json [0 "employee" "firstname"])
15:36	justin_smith	skoude: those still are not keywords
15:36	justin_smith	and the true option specifies it should make kewords
15:36	skoude	sorry pasted wrong line: ude (+i) 8:fn/#clojure (+cnt) Act: 1,2,3,4,5,6,11,13
15:37	justin_smith	(map #(get-in % [:employee :firstname]) parsed_json)
15:37	skoude	used get-in parsed_json [0 :employee :firstname])
15:38	skoude	and still get nil
15:38	justin_smith	what does it look like when you just enter parsed_json at the repl and see the contents?
15:38	skoude	it looks perfectly okay :)
15:38	justin_smith	skoude: can I see it?
15:39	skoude	justin_smith: I cannot paste it straight because it's a json from production system... But it starts like:
15:40	skoude	[
15:40	skoude	{
15:40	skoude	"employee": {
15:40	skoude	"firstname": "KAI",
15:40	skoude	"lastname": "KALLUNKI",
15:40	justin_smith	that's not parsed
15:40	justin_smith	that's json
15:40	skoude	I use println to print it, and I tough that just a clojures way to display it?
15:41	justin_smith	no, clojure will not display a real hash-map that way, use prn on it
15:41	justin_smith	I bet you will discover an unparsed string with prn
15:42	skoude	yeah prn displays: "[\n {\n \"employee\": {\n \"firstname\": \"KAI\",\n
15:42	justin_smith	yes, that is an unparsed string
15:43	skoude	hmm.. strange, because I still call: (def parsed_json( parse json))
15:44	justin_smith	is the json somehow double-encoded?
15:44	skoude	and parse is the function that has: (cheshire.core/parse-string s true)) -defined
15:46	skoude	aah now I got it.. the problem was this: (def json (pr-str(:body (http-get-request
15:46	skoude	it should be only: (def json (:body (http-get-request "http
15:48	skoude	now if I do prn it shows like: ({:employee {
16:03	skoude	Thanks for the help, it works perfectly now.. Just takes a little time to get used to clojure when coming from java
16:05	justin_smith	in general, don't trust println to show you what something actually is - use prn for less ambiguity
16:05	skoude	thanks for the tip, will do and use that
16:06	justin_smith	,(println {:a "0 :b 1"})
16:06	clojurebot	{:a 0 :b 1}\n
16:06	justin_smith	,(prn {:a "0 :b 1"})
16:06	clojurebot	{:a "0 :b 1"}\n
16:09	cloj_dev	whats wrong with this destructuring of a vector? http://pastebin.com/kbvDGxCj
16:24	Glenjamin	you have two arguments
16:24	Glenjamin	a map, then a vector
16:24	Glenjamin	but you're calling it with only a vector
16:25	ianhedoesit	also your map destructuring is not probably how you expect it to be.
16:26	ianhedoesit	if you try `(foo {0 :a 3 :b} [1 2 3 4])` you might see part of the problem.
16:26	ianhedoesit	I might recommend this helpful cheatsheat (https://gist.github.com/john2x/e1dca953548bfdfb9844) to compare - specifically https://gist.github.com/john2x/e1dca953548bfdfb9844#maps in this case.
16:27	justin_smith	,(let [{a 0 b 1} [:foo :bar]] [b a])
16:27	clojurebot	[:bar :foo]
16:27	jonathanj	hmm, how do i return a value from :post! in liberator?
16:28	justin_smith	ianhedoesit: the map syntax works fine for selecting indexes, see above
16:28	ianhedoesit	justin_smith: huh, interesting.
16:28	justin_smith	ianhedoesit: because vectors are associative by index
16:28	ianhedoesit	right, that makes sense.
16:28	irctc	Hello everyone.
16:28	justin_smith	it should fail with a list though
16:29	ianhedoesit	(it does)
16:29	irctc	I need help setting up reagent. I've been trying to figure it out but I just can't get it to work. Can somebody help me please?
16:29	ianhedoesit	I'm still going to assume that that wasn't what cloj_dev was trying to do, however. but maybe I'm wrong
16:30	justin_smith	irctc: do you have specific questions?
16:31	cloj_dev	lein new reagent
16:31	irctc	Yes. I put reagent in the dependencies in the project.clj file in my compojure project and Eclipse/counterclockwise plugin downloaded them all automatically for me. Now, how do I actually imlplement reagent in my web pages?
16:32	irctc	Do I put require [reagent.core :as r] in my views.clj file that renders my webpage and write my reagent code there or should I do something else?
16:33	justin_smith	irctc: first step is to set up clojurescript. reagent is not clojure (clj) it is clojurescript (cljs)
16:34	irctc	Ok, so in order for reagent to work I first need to download the dependencies for clojurescript?
16:34	justin_smith	irctc: next step is to have a cljs file that calls reagent.core/render-component, that sets up the rendering and you need to pass it an atom, changes to that atom will make new re-rendering of the page happen
16:34	justin_smith	irctc: your dep manager should do that for you, what you need to do is define a clojurescript build in your project
16:34	justin_smith	irctc: are you using leiningen?
16:35	irctc	Yes I am. leiningen and compojure for my web app.
16:35	irctc	And hiccup to render html5 components.
16:35	justin_smith	OK, so your first step is to check out lein-cljsbuild which you can use to set up clojurescript building
16:36	justin_smith	irctc: reagent is similar to hiccup (uses the same input syntax just about) but runs on the browser instead of your server
16:37	justin_smith	irctc: once you can get a basic hello-world cljs built and embedded into your html page and displaying when you load the page, you can move on to pulling in reagent
16:37	irctc	Ok, so I should set up lein-cljsbuild in my compojure app before I set up reagent?
16:37	cloj_dev	I'm trying to do something like this in clj http://pastebin.com/8mz7yZBT
16:37	cloj_dev	I know why its wrong but I can't figure out a good solution
16:38	justin_smith	irctc: right, reagent is a dep that cljsbuild will use, cljsbuild will make javascript from your clojurescript code, that the browser will run to create your page
16:39	irctc	Ok, great. Thanks for your guidance. I might have some more questions as I go along setting it up, but I'm gonig to try and set everything up now.
16:39	justin_smith	lein-cljsbuild and the reagent project both have good docs
16:40	cloj_dev	reagent is very easy to get going if you have lein
16:40	ianhedoesit	cloj_dev: are you trying to reverse a vector?
16:40	cloj_dev	sort of
16:41	ianhedoesit	or are you trying to stagger it backwards?
16:41	justin_smith	cloj_dev: that won't ever return - you need a conditional around recur
16:41	cloj_dev	I want (foo [1 2 3 4]) to be [2 1 4 3]
16:42	justin_smith	yeah, all you need is an if check for rest being empty, and just returning [b a] if rest is empty
16:42	justin_smith	well - you might need one or two other things, but that will be a start
16:42	jonathanj	any recommended HTTP client libraries?
16:43	jonathanj	i have no idea how to choose between the top few google results
16:43	justin_smith	jonathanj: I've found http-kit.client better than clj-http.client
16:43	jonathanj	justin_smith: why?
16:44	justin_smith	clj-http has some bad behaviors for poorly behaved endpoints in my experience, and wasn't nearly as nice for async stuff
16:44	justin_smith	for example timeout arguments seemingly not working as expected
16:44	cloj_dev	hm this function won't work
16:44	cloj_dev	http://pastebin.com/zhmDfU09
16:45	jonathanj	justin_smith: thanks, that sounds like a good recommendation
16:46	ianhedoesit	cloj_dev: what do you mean it won't work?
16:46	justin_smith	cloj_dev: do you expect your input to be individual items, or a sequence?
16:46	cloj_dev	a sequence I beliece
16:46	ianhedoesit	(foo 1 2 3 4) outputs "fun"
16:46	cloj_dev	like [1 2 3 4]
16:46	cloj_dev	oh
16:46	justin_smith	cloj_dev: because as of now it is looking for individual items, you need an extra pair of [] around the arg list
16:46	cloj_dev	ah okay
16:46	cloj_dev	good call
16:47	justin_smith	because the recursive version would be a pain without that
16:47	cloj_dev	thanks
16:57	irctc	I ran into a problem installing lein-cljsbuild. When I try to do lein cljsbuild auto like it says on the github page: https://github.com/emezeske/lein-cljsbuild I get an error.
16:57	irctc	lein cljsbuild auto Watching for changes before compiling ClojureScript... Exception in thread "main" java.io.FileNotFoundException: Could not locate cljs/util__init.class or cljs/util.clj on classpath: , compiling:(cljs/closure.clj:1:1)
16:58	irctc	I have no idea what that exception is all about.
16:58	justin_smith	irctc: can you share your project.clj on refheap?
17:00	irctc	Sure. Here it is: https://www.refheap.com/78c74e95513c162e1816f5cec
17:02	justin_smith	my first guess is an incompatibility between your clojurescript dep version and the lein-cljsbuild dep version
17:03	irctc	I just took the latest ones, like it says on their github pages.
17:03	cloj_dev_	hm, this gives an infinite loop http://pastebin.com/Bbk4rS9x
17:04	justin_smith	irctc: sure, but I don't think they can guarantee they are compatible with every cljs version - and the error you got looks like an incompatible version issue between cljs and cljsbuild
17:05	justin_smith	cloj_dev_: you are never modifying that "rest" argument - it will never change in that loop
17:05	justin_smith	so if it starts non-empty, it can never end up empty
17:05	cloj_dev_	ah right
17:06	irctc	Ok, I'm going back to 0.0-3190 Hopefully that works.
17:06	irctc	I tried 1.7.10 but that just produced the same error.
17:11	irctc	Ok if it works is it supposed to just say Watching for changes before compiling ClojureScript... and not give me back my prompt?
17:11	justin_smith	right
17:11	justin_smith	unless you provide the "once" arg
17:11	justin_smith	if you provide once it will exit after compiling
17:11	justin_smith	it will recompile if it sees new cljs files or edits to existing files
17:12	justin_smith	next step is to pull the generated js into your html so that it can run
17:12	irctc	Ok. And how do I now run my lein ring server command? Do I just leave this terminal window open and just run it from another one?
17:12	justin_smith	yeah, just run it in a different terminal
17:13	justin_smith	stepping away from the computer for a bit... I'll check in later
17:13	irctc	Oh, ok, so the next step for me is to actually create a clojurescript file and then incorporate it into my views.clj function that will generate my page?
17:13	irctc	Ok, thanks a lot for all your help. :)
17:13	jonathanj	justin_smith: how do i send a custom body with http-kit.client?
17:14	jonathanj	justin_smith: in this case i want to send a JSON body in a POST request
17:14	jonathanj	i see :form-params but it sounds like it's form-encoded?
17:15	jonathanj	apparently :body is an option, although the client docs <http://www.http-kit.org/client.html> fail to mention it
17:16	cloj_dev_	darn, still getting an infinite loop http://pastebin.com/ZmDbh6T5
17:21	ianhedoesit	cloj_dev_: you're still not doing anything with the rest of the list. if you change `[d c]` to `(println [d c])` you'll see what's happening
17:24	cloj_dev_	hm tried that
17:24	cloj_dev_	it did what I expected I think
17:26	oddcully	jonathanj: something like e.g. (es/post (str uri url-part) {:content-type :json :as :json :body (json/generate-string request)}))
17:52	irctc	In clojurescript in my project.clj where it says: :source-paths ["src-cljs"] where is that source path actually supposed to be?
17:52	irctc	Is it supposed to be on the same level as src or in the resources or in resources/public?
17:52	irctc	Or somewhere else?
18:05	justin_smith	irctc: that is the path from the project.clj to the top level of the sources
18:06	justin_smith	so yeah, same level as src
18:07	irctc	Thank you. :)
18:08	irctc	I actually compiled my clojurescript file into a main.js file, but I can't see the newly formed file in my eclipse editor.
18:09	irctc	Nevermind. I found it in my file system.
18:09	irctc	For some reason eclipse is just not showing it in the file structure.
18:11	irctc	So when I have multiple clojurescript files, will they all get compiled into this single main.js file?
18:16	justin_smith	the main.js file will make the rest of them load, but you don't get the single file without advanced compilation iirc
18:18	irctc	Ok.
18:18	irctc	And now when I call the main.js in my views.clj file, how do I point to the actual function that I want to display its results on my page?
18:19	irctc	I am using this to call the main.js file: (hic-p/include-js "/js/main.js")
18:19	justin_smith	your html should load your main.js and run some function in it
18:19	irctc	That doesn't happen.
18:19	justin_smith	irctc: it has to
18:20	justin_smith	irctc: there's no magic that makes a browser run clojurescript
18:20	justin_smith	there has to be a js file, and there has to be some html file that makes the browser load the js file
18:20	justin_smith	your cljs is just there to be compiled and turned into that js
18:20	irctc	I put the call to the main.js in my hed generating function which works as it generates all other html elements, but it doesn't render the simple "Hello world text" from my clojurescript file.
18:21	irctc	Head generating function*
18:21	justin_smith	OK sure, you can generate the html
18:22	justin_smith	point is, the entry point is the html which the browser understands, then you set up your rendering to happen when your main method in cljs gets called
18:22	justin_smith	eventually, that should be a call to reagent
18:22	justin_smith	once you establish you can get your cljs to load and run
18:23	irctc	So do I call the cljs function from my views.clj file just like I would with any other function?
18:23	justin_smith	how would your clj code make anything happen in the browser?
18:24	justin_smith	I mean it can, but you need some mechanism of communication...
18:26	irctc	Well this is what I am confused about. I am calling the main.js file and I am assuming that it loads because I loaded a plain javascript file the same way. Next I want it to display the words "Hello world" on my page that the clojurescript function (hello) makes, but when I open my browser window that doesn't write anything on my page.
18:28	rhg135	You have to call that function
18:28	justin_smith	irctc: right, this is the part where reagent comes in. First verify your cljs runs (eg. you can run (.alert js/window "hello") inside your cljs and verify you get the popup), after that to get rendering happening, that is when you start using reagent
18:28	irctc	If the main.js file makes the rest of the clojurescript files load, then how do I access my hello function in example.cljs and load the results of executing that function on my page?
18:28	justin_smith	rhg135: well, just because some function returned "hello world" isn't enough to make the browser show the message
18:29	justin_smith	irctc: you need a client side renderer, that's what reagent does
18:29	justin_smith	irctc: or you can play with the dom, or with jquery, but right now those would both be wastes of your time
18:29	irctc	Ok, I'm going to verify that it works first with that alert message.
18:30	rhg135	Oh, I misread, justin_smith, right
18:30	justin_smith	cool, and if that works, you can start hooking up reagent, then things get fun :)
18:30	irctc	I can't wait. Keeping my fingers crossed. :)
18:32	irctc	It worked!!! :D
18:33	irctc	I don't know I've ever been so happy to see an alert message before. lol :D
18:34	justin_smith	irctc: cool, next step, now that you know you can compile cljs and make it run, is make sure you have reagent in your deps, and call reagent/render-component with the apropriate args
18:34	justin_smith	irctc: spend some time with the reagent docs, they are really good and thorough, and you should be able to follow along with any examples now that you have this much set up
18:35	justin_smith	irctc: reagent/render-component will watch an atom, and every time the atom changes, it will update the contents of the page (that's the big picture, of course there are lots of details to set up)
18:36	irctc	Thanks so much! :)
18:36	irctc	Btw, when calling reagent/render-component, do I call that in my views.clj file or in the cljs file?
18:37	justin_smith	in the cljs
18:37	irctc	And when I require reagent do I need to require it in my views.clj as well or just in my cljs file?
18:37	justin_smith	since it needs to run in the browser
18:37	justin_smith	irctc: what does views.clj do?
18:37	irctc	Generates my pages.
18:37	irctc	Part of my compojure framework.
18:38	justin_smith	yeah, this is actually an alternate way to generate pages - because compojure doesn't run in the browser
18:38	justin_smith	the concept is that instead of making the page on the server and sending to the client, the cljs code constructs the page on the client side, and just asks the server for data - views.clj would be used to make handlers that send data the frontend needs
18:39	irctc	Ok, so in my cljs file, right after I write my function I'll call (reagent/render-component) and that will watch over all of my atoms, and then I can just compile it and load it the same way I did with this previous alert message?
18:41	irctc	Ok, cool, because then I can use the client to make some async calls to the server to avoid blocking the user.
18:41	justin_smith	something like that - add the render-component to the main cljs function, and what the reagent docs will show is how to hook up the atoms to manage the page state
18:41	irctc	That's the whole reason why I'm using cljs and trying to use reagent in the first place. :)
18:41	justin_smith	and yeah, make async calls to the server (calling functions in views.clj) that load the data you need to load into the page
18:41	justin_smith	right
18:42	irctc	Cool, thanks. :)
18:42	irctc	I am using this for the reagent tutorial https://reagent-project.github.io/ Hopefully I'll be successful in setting it all up.
18:43	justin_smith	yup, that's the one
18:44	justin_smith	irctc: if you scroll down a bit, it shows how to set up the cljs namespace and call render-component
18:46	irctc	Cool. Thanks. :)
18:50	irctc	I copied and pasted the entire first example that mentions reagent/render-component except for the name of the namespace just to test it out but I can't get anything to display on my page.
18:51	justin_smith	OK, I bet you will find compilation errors in your cljsbuild window, or js errors in the browser console
18:52	irctc	Compiling ClojureScript... Compiling "resources/public/js/main.js" from ["srccljs"]... Successfully compiled "resources/public/js/main.js" in 7.157 seconds.
18:52	irctc	Everything seems allright.
18:52	irctc	I even restarted the server.
18:52	justin_smith	ok, how about the browser js console
18:53	irctc	You were right. There is something in the browser js console: mutating the [[Prototype]] of an object will cause your code to run very slowly; instead create the object with the correct initial [[Prototype]] value using Object.create
18:53	irctc	Although I don't know what that means.
18:54	justin_smith	that's not an error though
18:56	irctc	I know, but that's all I get and I don't know why I'm not getting any output on my page.
18:57	justin_smith	what do you see when you inspect the page (via right-click)
18:58	irctc	I see it loads the main.js file: <script src="/js/main.js" type="text/javascript"></script>
18:58	justin_smith	is there anything in the page body?
18:59	irctc	Other than that there's nothing else there.
18:59	justin_smith	irctc: in the r/render-component call, what is the last arg?
19:00	irctc	(.-body js/document)
19:00	justin_smith	OK, so you tell reagent to render to the document body, but when you inspect the page there is no body?
19:01	irctc	Exactly.
19:02	justin_smith	how about making sure there is a body to the page so that it can render?
19:02	irctc	Well I mean there is a body, there's jut no content in it.
19:02	justin_smith	I have to go
19:03	irctc	Ok, thanks for all your help.
19:03	irctc	I really appreciate it.
19:03	irctc	:)
19:30	pyon	I want to make two macros, `deffoo` and `defbar`. `deffoo` creates a bunch of stuff, and the last thing it creates is a variable foo: (`def ~name { :meta ... :data ... :goes ... :here ... }). What I want `defbar` to do is (0) create a foo, (1) some bar-specific work, (2) insert more metadata into the metadata object.
19:31	pyon	In a language like Common Lisp, I'd just create the metadata object from `deffoo` and then mutate it from `defbar`.
19:31	pyon	What would a more idiomatic Clojure solution, without mutability, look like?
19:49	pyon	Is there any function like `map`, but which discards the result of applying the function to the list's elements?
19:50	pyon	Basically something like (keep func seq), except (func elem) always returns nil.
19:51	rhg135	run!
19:52	rhg135	in 1.8 of course
19:52	pyon	rhg135: Oh, nice, thanks!
19:52	rhg135	1.7*
19:52	pyon	I'm using 1.7.0, which comes with my package manager, and run! works.
19:52	rhg135	ok good
20:01	pyon	How do I check whether a symbol refers to an existing variable.
20:01	pyon	?
20:03	rhg135	resolve
20:04	rhg135	,(do (resolve 'x2) (resolve '+))
20:04	clojurebot	#'clojure.core/+
20:05	rhg135	uhh, the former returns nil
20:05	justin_smith	,(resolve 'to-quit-smoking)
20:05	clojurebot	nil
20:06	rhg135	,(resolve 'to-lose-weight)
20:06	clojurebot	nil
20:06	pyon	Ah!
20:06	justin_smith	haha
20:06	pyon	lol
20:06	pyon	rhg135: ty!
20:06	rhg135	not in cljs though
20:07	rhg135	in cljs you need to intrude on the compiler
20:07	justin_smith	rhg135: can't you do like (aget js/window "your-ns/your-var") or something like that?
20:08	rhg135	that assumes you want to hardcode the ns name. so forget DRY
20:08	justin_smith	rhg135: aget takes a string as an arg...
20:09	rhg135	cljs lacks getting the name at runtime, at least back in 201.r4
20:09	pyon	Is there any way to tell the Clojure REPL “forget every single thing I've defined before”, without restarting it?
20:10	justin_smith	rhg135: it's an array operation, you are treating the object as an array and looking up a key in it
20:10	rhg135	,(run! (partial ns-unmap ns) (ns-publics *ns))
20:10	clojurebot	#error {\n :cause "Unable to resolve symbol: ns in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: ns in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: *ns in this conte...
20:11	rhg135	err, something lise that
20:11	justin_smith	run takes a function and a collection
20:11	justin_smith	err, run! does I mean
20:12	pyon	:-O
20:13	rhg135	justin_smith: I mean you can't know the actual ns name at runtime without hardcoding at compile time. unless dnolen endowed it by now
20:14	Bronsa	justin_smith: that won't work with advanced compilation
20:14	justin_smith	Bronsa: ahh, good point
20:14	Bronsa	I don't think there's a good way to do resolve in cljs
20:14	justin_smith	rhg135: the ns at runtime is 'user anyway, aside from dev time shenanigans
20:14	rhg135	oh and that. thx Bronsa
20:15	rhg135	'cljs.user but yeah
20:16	rhg135	you can do it at compile time via the api
20:16	rhg135	I think
20:32	pyon	How could I make a macro for asserting multiple things in a single place?
20:32	pyon	(multi-assert test1 "failed test1" test2 "failed test2" ...)
20:51	rhg135	,(defmacro multi-assert [& pairs] `(do ~@(for [[t msg] (partition 2 pairs)] `(assert ~t ~msg))))
20:52	clojurebot	#'sandbox/multi-assert
20:52	justin_smith	why would this need to be a macro?
20:52	rhg135	ah. yeah he's right
20:54	justin_smith	well, because assert is a macro I guess...
20:54	rhg135	maybe lazyness, aka if a certain test is very expensive
20:54	justin_smith	though we can still do this without another macro
20:55	justin_smith	,(map #(assert (first %) (second %)) [[true "ok"] [false "oops"]])
20:55	clojurebot	#<AssertionError java.lang.AssertionError: Assert failed: oops\n(first p1__68#)>
21:32	pyon	rhg135: justin_smith: ah, thanks!
21:36	pyon	https://www.refheap.com/111249 -- here, in constructor-metadata, how do I quote the name?
21:37	pyon	If I do 'name , it gives me (quote name), but what I actually want to quote is the value of the variable name.
21:38	pyon	To be clearer - how do I quote the value of a variable?
21:38	justin_smith	either the caller needs to quote name, or constructor-metadata needs to be a macro
21:39	justin_smith	you can't get the quoted version of an argument in a function
21:39	pyon	Oh.
21:39	pyon	But, if I make it a macro, then I can't map over it in defpolynomial. Mmm...
21:40	justin_smith	sounds like the caller needs to quote the names
21:41	pyon	I can think of ways to do it, but not in syntactically pleasing ways. :-\|
21:47	cloj_dev	I'm still working on this problem : http://pastebin.com/9mBXSgxm
21:47	cloj_dev	anyone have a suggestion?
21:48	justin_smith	cloj_dev: next never changes
21:48	cloj_dev	huh
21:49	cloj_dev	ah I see that now
21:49	justin_smith	,((fn [[a b & rest]] (if (empty? rest) [b a] (concat [b a] (recur rest)))) [1 2 3 4 5 6])
21:49	clojurebot	#error {\n :cause "Can only recur from tail position"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.UnsupportedOperationException: Can only recur from tail position, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyzeSeq "Compiler.java" 6891]}\n {:type java.lang.UnsupportedOperationException\n :message "Can only recur from tail position"\n ...
21:49	justin_smith	hrmph
21:50	justin_smith	,((fn flipper [[a b & rest]] (if (empty? rest) [b a] (concat [b a] (flipper rest)))) [1 2 3 4 5 6])
21:50	clojurebot	(2 1 4 3 6 ...)
21:50	justin_smith	it's still bad if the arg count is odd though
21:50	cloj_dev	oh nice
21:51	cloj_dev	I thought you can't have recursive functions that don't call recur
21:51	justin_smith	you can, they just don't get optimized
21:51	justin_smith	but with laziness (concat) that is less of an issue
21:51	cloj_dev	oh
21:54	justin_smith	cloj_dev: it's normal for lazy functions to use direct self-call instead of recur
21:54	cloj_dev	interesting
21:54	cloj_dev	how do you know if a function is lazy?
21:54	cloj_dev	is it self evident, or documented?
21:54	cloj_dev	or you just know by rote?
21:54	justin_smith	there's basically one lazy class: clojure.lang.LazySeq
21:55	justin_smith	lazy functions are the ones that use that class directly (via the lazy-seq function) or call others that use it (like map or concat or filter)
21:55	justin_smith	functions should document if they are lazy
21:55	cloj_dev	right cool
21:56	rarebreed	there's also the delay macro
21:56	justin_smith	that's not lazy
21:56	rarebreed	which will delay evaluation of a function
21:57	justin_smith	yes, but it's not considered lazy by clojure's standards - with a lazy seq accessing the value can force execution, with delay you need to deref
21:57	justin_smith	I'd group delay closer to future and promise
21:57	justin_smith	in terms of spheres of functionality
21:58	rarebreed	yes, if you use delay, you have to force on the result or deref it
21:58	justin_smith	things that could give you a value, but you may need to wait, and you can check exlicitly whether they are realized
22:01	pyon	Ok, just to be sure. What's the simplest way to make a macro quoteall, such that (quoteall foo bar qux) expands to ['foo 'bar 'qux] ?
22:05	pyon	,(defmacro unusable [foo] (class foo)) (unusable 'foo) ; why?
22:05	clojurebot	#'sandbox/unusable
22:06	pyon	,(unusable 'foo) ; why?
22:06	clojurebot	clojure.lang.Cons
22:06	pyon	How is this a cons cell?
22:12	justin_smith	,''foo
22:12	clojurebot	(quote foo)
22:12	justin_smith	that's why
22:13	gfredericks	,(defmacro quoteall [& args] (mapv #(list 'quote %) args))
22:13	clojurebot	#'sandbox/quoteall
22:13	gfredericks	,(quoteall foo bar qux)
22:13	clojurebot	[foo bar qux]
22:13	justin_smith	~macros
22:13	clojurebot	macros are just a game with symbols
22:13	gfredericks	,(defmacro quoteall [& args] `'[~@args])
22:13	clojurebot	#'sandbox/quoteall
22:13	gfredericks	,(quoteall foo bar qux)
22:13	clojurebot	[foo bar qux]
22:13	gfredericks	haha I just wrote `'[~@args] #yolo
22:14	pyon	justin_smith: Ah!
22:14	justin_smith	gfredericks: you could have gotten a #reader-macro in there if you tried
22:15	justin_smith	and #yolo doesn't count :P
22:15	pyon	,(class (first (quoteall (foo bar qux)))
22:15	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading>
22:15	pyon	oops
22:15	pyon	,(class (first (quoteall (foo bar qux))))
22:15	clojurebot	#error {\n :cause "Unable to resolve symbol: quoteall in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: quoteall in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: quoteal...
22:15	pyon	I'm dumb.
22:15	pyon	,(class (first (quoteall foo bar qux))
22:15	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading>
22:15	pyon	Never mind.
22:16	gfredericks	,(class (first (quoteall foo bar qux)))
22:16	clojurebot	#error {\n :cause "Unable to resolve symbol: quoteall in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: quoteall in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: quoteal...
22:16	gfredericks	,(defmacro quoteall [& args] `'[~@args])
22:16	clojurebot	#'sandbox/quoteall
22:16	gfredericks	,(class (first (quoteall foo bar qux)))
22:16	clojurebot	clojure.lang.Symbol
22:16	pyon	Ah!
22:16	pyon	What exactly does `' do?
22:16	gfredericks	clojurebot is forgetful
22:17	pyon	Heh, makes sense. :-)
22:17	justin_smith	pyon: it wraps quote in syntax-quote, and it is usually a sign of mischeaf
22:17	gfredericks	,(defmacro quoteall [& args] (list 'quote (vec args)))
22:17	clojurebot	#'sandbox/quoteall
22:17	gfredericks	,(quoteall foo bar qux)
22:17	clojurebot	[foo bar qux]
22:17	gfredericks	pyon: ^ that oughta be a much simpler definition to understand
22:17	pyon	Ah :-)
22:21	amalloy	,'`'[~@args]
22:21	clojurebot	(clojure.core/seq (clojure.core/concat (clojure.core/list (quote quote)) (clojure.core/list (clojure.core/apply clojure.core/vector (clojure.core/seq (clojure.core/concat args))))))
22:21	amalloy	^ that oughta be a less simple explanation to understand
22:21	justin_smith	lol
22:23	OrbitalKitten	hi guys, do you know if there's a way to generate the members in deftype/defrecord without going into eval hell?
22:23	gfredericks	,(inc amalloy)
22:23	clojurebot	#error {\n :cause "Unable to resolve symbol: amalloy in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: amalloy in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: amalloy i...
22:23	gfredericks	(inc amalloy)
22:23	gfredericks	OrbitalKitten: I think defrecord has a way
22:24	gfredericks	,(defrecord OrbitalKitty [weight velocity])
22:24	clojurebot	sandbox.OrbitalKitty
22:24	gfredericks	,(OrbitalKitty/getBasis)
22:24	clojurebot	[weight velocity]
22:24	OrbitalKitten	ooh by member I mean implementations of the methods
22:24	gfredericks	ooooh you want to generate implementations of protocol programmatically
22:24	OrbitalKitten	yeah
22:24	gfredericks	,(doc extend)
22:24	clojurebot	"([atype & proto+mmaps]); Implementations of protocol methods can be provided using the extend construct: (extend AType AProtocol {:foo an-existing-fn :bar (fn [a b] ...) :baz (fn ([a]...) ([a b] ...)...)} BProtocol {...} ...) extend takes a type/class (or interface, see below), and one or more protocol + method map pairs. It will extend the polymorphism of the protocol's methods to call the suppl...
22:25	gfredericks	it's surprisingly easy :)
22:25	OrbitalKitten	thanks, I will look at this
22:30	OrbitalKitten	hmm looks like it only works with Protocols, and i'm implementing a huge ass interface
22:30	OrbitalKitten	basically it's a visitor that has a ton of visit/endVisit methods
22:31	OrbitalKitten	I'm using reflection to find out what I need to implement and generate the corresponding methods
22:31	OrbitalKitten	heh I'll stick with the evil code for now
23:12	pyon	Does Clojure have anything like Common Lisp's keyword arguments?
23:33	TEttinger	pyon: I don't know CL, but we do have keyword destructuring for maps as args. I think it's frowned upon
23:34	TEttinger	~kwargs
23:34	clojurebot	kwargs are a bad idea.
23:34	TEttinger	~kwargs
23:34	clojurebot	kwargs are a bad idea.
23:34	TEttinger	hm
23:34	pyon	Ah!
23:34	TEttinger	there's another one
23:34	TEttinger	it doesn't play well with other clojure stuff though
23:34	TEttinger	since you can't make a partial fn of it, for instance
23:35	pyon	Makes sense.
23:35	TEttinger	it technically takes a single arg, a map that may contain lots of keywords and values, so partial would need a whole new map
23:37	pyon	Can I understand a macro as a normal function, except its arguments are passed all quoted?
23:39	amalloy	pyon: yes, except that also its result is expanded into the function calling it, rather than being returned as a value
23:39	pyon	Ah!
23:40	pyon	So, in principle, I would write everything in terms of normal functions, and use a single macro `macroize`, that lets me use other existing functions as macros, right?
23:40	pyon	s/would/could/
23:41	TEttinger	usually you should prefer fns because macros can't be passed around as first-class values
23:41	TEttinger	,(reduce and [true true true])
23:41	pyon	Yeah. I've been running into that a lot.
23:41	clojurebot	#error {\n :cause "Can't take value of a macro: #'clojure.core/and"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Can't take value of a macro: #'clojure.core/and, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Can't take value of a macro: #'clojure....
23:50	pyon	So, I could use (defmacro macroize [function & arguments] `(~function ~@(map #(list 'quote %1) arguments))), and define all other macros in terms of normal functions and macroize, right?
23:50	pyon	Or, is this a bad idea for some reason?
23:51	amalloy	it's not clear how you intend macroize to work but i'm fairly sure it won't work that way
23:58	pyon	Can I “delay a ~”?
23:58	pyon	I'm trying to write a macro that produces a macro.
23:59	spradnyesh	can someone please explain this?
23:59	spradnyesh	(= (double 1.0) (float 1.0)) => true
23:59	spradnyesh	and (= (cast java.lang.Double 1.0) (cast java.lang.Float (float 1.0))) => true
23:59	spradnyesh	but (= (:close (second data)) (:close (second rslt))) => false, where
23:59	spradnyesh	(:close (second data)) => 95.2
23:59	spradnyesh	(:close (second rslt)) => 95.2
23:59	spradnyesh	(type (:close (second rslt))) => java.lang.Float
23:59	spradnyesh	(type (:close (second data))) => java.lang.Double