#clojure – 2014-04-19

2014-04-19

00:01	bodie_	this is probably a silly question, but let's say I want to work on a library project (say a DigitalOcean API client), which I'm testing using a project with a -main. how do I include the lib project?
00:01	bodie_	(using Lein.)
00:01	bodie_	or uh, where do I need to start looking to understand this, that is?
00:02	bodie_	I get how to include external deps using lein... would I want to uberjar the subproject and include the jar as a resource?
00:09	owl-v-	how do i chanve :else part so that it runs without error? https://www.refheap.com/77460
00:10	owl-v-	putting elements of v into vv one by one in a recursive-function
00:15	owl-v-	solved my problem. but output is not correct
00:16	owl-v-	how do i get output '(1 2 3) from this code? https://www.refheap.com/77461
00:20	owl-v-	solved!
00:21	owl-v-	https://www.refheap.com/77462
00:21	owl-v-	solved :)
01:26	teltubs	anybody got lighttables to work on linux ?
01:29	teltubs	exit
02:06	mr-foobar	hi all, need some help I'm followinghttps://github.com/clojure/clojurescript/wiki/Quick-Start but I can't execute `:target: nodejs` code because it is outputting goog... code in the output
03:06	yedi	yesssssssssssss
03:18	owl-v-	why is it hard to debug in clojure?
03:20	oskarkv	owl-v- is it hard? :P
03:22	owl-v-	it's much harder to debug in clojure than in python.
03:22	owl-v-	error prompt is not useful.
03:23	owl-v-	it's not telling what's wrong about my code
03:24	owl-v-	it just points to some line
03:24	oskarkv	owl-v- you mean stacktraces?
03:25	owl-v-	m.. maybe.. yes.
03:25	oskarkv	show me one :p
03:25	oskarkv	example
03:27	owl-v-	https://www.refheap.com/77486
03:28	oskarkv	owl-v- it does tell you what's wrong! "wrong number of args (0)..."
03:29	owl-v-	I can't tell where the # of args don't match
03:29	oskarkv	paste the code too and i'll help you :p
03:30	owl-v-	thanks T.T https://www.refheap.com/77487
03:30	amalloy	i predict that bfs.invoke(add.clj:45) contains a call to (apply max xs), where xs is empty
03:31	owl-v-	it happened when i added (digit3) condition in bfsfun
03:32	oskarkv	owl-v- it's probably what amalloy said
03:33	owl-v-	hm... then (apply max) can't use on empty list/vector...
03:33	amalloy	well, what is the max of an empty list?
03:33	owl-v-	,(apply max '())
03:33	clojurebot	#<ArityException clojure.lang.ArityException: Wrong number of args (0) passed to: core/max>
03:33	owl-v-	!
03:33	amalloy	owl-v-: should it invent a number? how can you get the max?
03:34	owl-v-	hehe empty list
03:35	oskarkv	owl-v- Once you get more used to Clojure debugging is not that hard. Don't be discouraged! :p
03:36	owl-v-	thanks for your encouragement!
03:37	amalloy	it's true. i only needed to look at that stacktrace for about five seconds before i predicted the mistake. you can get there too
03:39	owl-v-	funny, the stack points to (for) where the actual problem happens on (max)
03:39	owl-v-	or does it points to the argument of (max)?
03:41	amalloy	maybe you have the line numbers wrong. it should be pointing to apply
04:40	owl-v-	i have no idea where (recur) goes to in this case. https://www.refheap.com/77488
04:41	owl-v-	i thought it (recur) to bfsfun with (pup u) but it shows the same value after iteration
04:41	owl-v-	something fishy about (let)...
04:50	owl-v-	the value is not equals to expected value: https://www.refheap.com/77494
04:52	owl-v-	is the problem from (let) or (recur)?: https://www.refheap.com/77494
05:04	owl-v-	in a recursive-function, (let) value don't change after (recur) ?
05:10	owl-v-	omg! i simply replaced (pop) to (rest) and it worked!: https://www.refheap.com/77494
05:11	owl-v-	what was i thinking ...
05:41	owl-v-	where would u recommend clojure over python? data-base? ai? embedded-system?
05:52	penthief	Is there a way to stop a blocking socket read without taking down the whole process when in a cider repl?
05:53	coventry`	owl-v-: I've used python since the 90s. I would say python still has the edge in scientific programming and visualization, and may be superior in other areas where it has very mature libraries. However as general programming languages, clojure's thoughtful state management policies give it a clear edge.
06:28	owl-v-	coventry`: thanks.
06:52	owl-v-	still.. lips way is un-natural to me, but it's interesting as hell :-)
06:57	owl-v-	(question "can clojure have an infinite loop?")
06:58	owl-v-	(question "will clojure eat up all memory if (recur) used for infinite loop?")
07:03	AimHere	owl-v-, yes. no.
07:03	AimHere	At least, as far as I'm aware
07:04	owl-v-	cool!
07:05	AimHere	Use this knowledge for good, not evil
07:07	owl-v-	um... (def good ..) (def evil ..)
07:11	owl-v-	how do I define good and evil?
07:12	owl-v-	is it global (def) or local (def)?
07:13	owl-v-	it would be interesting to see implementation of good/evil in artificial-intelligence .
09:28	owl-v-	is there way to add map? {:a 1} + {:a 2} = {:a 3}
09:29	owl-v-	is there way to add map? {:a 1, :b 2} + {:a 2, :b 1} = {:a 3, :b 3}
09:29	jkj_	interesting... is there map vals or so
09:29	Bronsa	,(merge-with + {:a 1 :b 2} {:a 2 :b 1})
09:29	clojurebot	{:b 3, :a 3}
09:30	owl-v-	lol. i don't see (merge-with) in this website: http://clojuredocs.org/clojure_core/1.2.0/clojure.core/map
09:32	AimHere	That's because that's an (outdated) page on the 'map' function, not the map data structure
09:32	owl-v-	and this too: http://clojuredocs.org/clojure_core/clojure.core/map
09:32	jkj_	clj-kitchensink might do it: https://github.com/puppetlabs/clj-kitchensink/blob/master/src/puppetlabs/kitchensink/core.clj#LC315
09:33	jkj_	deeply... merge-with is nice if no deep required
09:33	Bronsa	owl-v-: map is a function that operates on seqs, it has nothing to do with merge-with
09:33	AimHere	I like the clojure cheatsheet meself. Try looking up the section 'Maps' in in here: http://clojure.org/cheatsheet
09:34	owl-v-	hm... actually... it does not work on "tryclj.com"
09:34	owl-v-	(merge-with {:a 1} {:a 1} ) -> {:a 1}
09:35	owl-v-	ops! forgot to put "+"
09:38	owl-v-	so.. clojuredocs.org is way out dated.
09:38	owl-v-	it's updated up to 1.3.0
09:54	cddr	What's the easiest way in clojure to override part of an interface (e.g. override the `write` method of an output stream)
09:59	cddr	Ah, maybe proxy
10:00	malyn	cddr: I use this for such questions -- https://github.com/cemerick/clojure-type-selection-flowchart
10:01	owl-v-	what's 2^3=8 in clojure?
10:03	cddr	Hm, I think the flowchart leads me to reify. What I want is a writer that instead of writing to the console, collects messages into an array so they can be verified in a test
10:05	malyn	cddr: Isn't that just with-out-str?
10:05	malyn	&(doc with-out-str)
10:05	lazybot	⇒ "Macro ([& body]); Evaluates exprs in a context in which out is bound to a fresh StringWriter. Returns the string created by any nested printing calls."
10:06	cddr	Maybe that would work
10:11	owl-v-	clojure needs MATH!
10:12	owl-v-	don't tell me to loop
10:12	owl-v-	because that's like looping (+) to do (*)
10:21	cddr	malyn: Thanks for being my duck. I was totally overthinking it
10:21	owl-v-	still can't find math
10:25	owl-v-	:-(
10:27	owl-v-	where is scientific mind in clojure?
10:29	owl-v-	T.T
10:39	owl-v-	oh, clojure. give me MATH
10:40	owl-v-	clojure, let me do (= 8 (^ 2 3))
10:41	weavejester	owl-v-: (Math/pow 2 3)
10:41	invasmani	owl-v-: https://github.com/clojure/math.numeric-tower
10:43	owl-v-	weavejester: OMG!!!! big.THANK.u
10:44	owl-v-	i was going to give up and use python to solve math problem.
10:44	owl-v-	I had to use (int (Math/pow 2 3) to get integer.
12:13	bbloom	i really want an auto-splicing sequence data type
12:13	bbloom	(conj [1 2 3] (splice 4 5)) => [1 2 3 4 5]
12:17	bbloom	conj is already polymorphic for maps additions:
12:17	bbloom	,(conj {} nil [:x 1] {:y 2 :z 3})
12:17	clojurebot	{:z 3, :y 2, :x 1}
12:17	bbloom	need a non-polymorphic conj for vectors, something like "append"
12:18	bbloom	then conj can auto splice :-)
12:18	jkj	concat too slow?
12:21	owl-v-	bbloom: what's (splice 4 5) ?
12:23	bbloom	owl-v-: some thing i just made up
12:24	bbloom	jkj: slow is not the issue, i want to be able to move the decision to concatenate from the receiver of the value to the generator of the value
13:08	owl-v-	bbloom: how does (splice) works?
13:08	owl-v-	bbloom: how does (splice) work?
13:16	TravisD	I am implementing some experiments where the goal is to compare different algorithms for solving the same problem. Right now, I am planning to make a protocol that defines the interface each algorithm should have, so that I can write the rest of the experiment code against these protocols. Is this abuse?
13:21	apricots	If I type “lein cljsbuild auto &” to background the command in my shell, instead of backgrounding, it replies with “[1]+ Stopped lein cljsbuild auto”. It’s a minor annoyance, but is there a way to around that?
13:28	Jaood_	apricots: no idea, maybe you should try screen/tmux
13:30	ismaelvc	Hello everyone! I'm reading the clojure introduction at clojure-doc, I'm currently on "namespaces", and after reading about (require '[clojure.string :as str]), I tried to do (require '[javax.swing.JOptionPane :as Pane]), but I get: user=> (require '[javax.swing.JOptionPane/showMessageDialog :as Diag]) FileNotFoundException Could not locate showMessageDialog__init.class or showMessageDialog.clj on classpath: clojure.lang.RT.
13:30	ismaelvc	I only know Python, I'm trying to do something analogous to this in Jython: >>> from javax.swing.JOptionPane import showMessageDialog as Diag >>> Diag(None, "Hello Jython!")
13:31	JohnJ	noob here, I have this result from a function => ({: 1}) , what is this? A map with a empty string as key?
13:32	JohnJ	ismaelvc: I would say you need a require in the project.clj, do you have it?
13:32	dbasch	apricots: lein cljsbuild auto < /dev/null &
13:32	mpenet	,(import 'javax.swing.JOptionPane)
13:32	clojurebot	#<ExceptionInInitializerError java.lang.ExceptionInInitializerError>
13:33	dbasch	apricots: and you probably want to redirect the output to a file
13:33	mpenet	ismaelvc: meh should work, in short java stuff gets imported with "import" not require
13:34	mpenet	,(import '(javax.swing JOptionPane))
13:34	clojurebot	#<NoClassDefFoundError java.lang.NoClassDefFoundError: Could not initialize class javax.swing.JOptionPane>
13:37	mpenet	ismaelvc: then (JOptionPane/showMessageDialog nil "meh")
13:39	r00k	Hi all, I'm trying to insert a record in a postgres db using korma. I'm calling (insert clients (values {:date_of_birth "1980-01-01"})). This gives me "ERROR: column "date_of_birth" is of type date but expression is of type character varying" from QueryExecutorImpl.java. This seems an odd error since pg happily accepts strings for that column in the psql console. Any thoughts?
13:40	r00k	(I realize that I'm touching pg and java here, and this might not be a clojure thing at all.)
13:44	ismaelvc	<JohnJ> No I don't have a require in project.clj, I just started using the REPL
13:45	zamaterian	r00k, try converting your string to a java.sql.Date
13:46	ismaelvc	<mpnet> Thanks that is closer to what I wnated, but still how can I make an alias for showMessageDialog?
13:49	r00k	zamaterian: Thanks! Giving that a shot.
13:49	ismaelvc	I want to write just (Diag nil "meh") instead of (JOptionPane/showMessageDialog nil "meh")
13:52	JohnJ	guys which data structure is this: {: 1} ?
13:52	bhauman	ismaelvc: (defn diag [a b] (JOptionPane/showMessageDialog a b))
13:54	r00k	zamaterian: So when I convert my string to a java.sql.Date, it includes a time portion, so it's certain day at midnight at UTC. When I insert that in pg, it's storing it an one day prior, presumably because I'm in -5 UTC. Think I should specify a time zone on the java.sql.Date? I'd really rather avoid timezones entirely.
13:54	r00k	s/an one/as one
14:03	ismaelvc	bhauman: Thanks that works as I intended! So there's no alias when importing, only while requiring, right?
14:04	owl-v-	oh.. man. how do i sum prime numbers in between 1 and 2000000 ?
14:04	owl-v-	i'm using brute force
14:04	bhauman	ismaelvc: you can only alias the namespace not the definitions (functions) within the space
14:05	owl-v-	it's taking a lot of time...
14:05	honza	is there a list of built-in clojure functions that return chunked seqs?
14:05	owl-v-	and it's still calculating...
14:06	zamaterian	r00k, I would look into using clj-time
14:07	zamaterian	r00k, https://github.com/clj-time/clj-time
14:09	r00k	zamaterian: Thanks, but I actually have been using that. Still having issues. I'll keep struggilng.
14:17	ismaelvc	bhauman: Then what would be the advantege of first importing and then defining a function, I think it is better and less code just in one go like this: user=> (defn dialog [message] (javax.swing.JOptionPane/showMessageDialog nil message)) #'user/dialog user=> (dialog "Hello Clojure!") nil. Is there any drawback from doing it this way? Thanks so much! :)
14:19	bhauman	ismaelvc: there are no drawbacks that I see to either, case by case
14:21	creese	I'm wondering what is best practice for app-level configs in a leiningen project. I want to store info like database-url, redis, etc. but somewhere I can keep out of version control. Is the ability built-in?
14:21	mpenet	creese: https://github.com/weavejester/environ is nice for this kind of things
14:22	mpenet	or just read an edn file that you pass as argument to your app -main
14:22	mpenet	many ways to do this (yaml comes to mind too)
14:24	bhauman	creese: look at the functions: slurp and read-string , it’s pretty tirvial to read a clojure data structure from a file
14:25	mpenet	http://clojure.github.io/clojure/clojure.edn-api.html this read-string, not the one from core preferably...
14:25	mpenet	,(doc read-string)
14:25	clojurebot	"([s]); Reads one object from the string s. Note that read-string can execute code (controlled by read-eval), and as such should be used only with trusted sources. For data structure interop use clojure.edn/read-string"
14:27	creese	mpenet: I've used yaml before, seems like edn is a better choice here
14:28	bhauman	mpenet: ok got it
14:29	bhauman	don’t know why I ignore clojurebot so much, he seems to know what he’s talking about
14:29	mpenet	I like edn better too, but if you need to expose it to ppl who are not familiar yaml (or something else) can be a better choice
14:39	sm0ke	i can see a cljs/core.js in the :output-dir, but still my browser complains "goog.require could not find: cljs.core "
14:40	sm0ke	i was just trying to get the brepl working
14:40	sm0ke	https://github.com/cemerick/piggieback#browser-repl
14:44	amalloy	honza: well, the main ones are like...range, and seq on a vector. then there are the functions that preserve chunkiness, and the ones of those that i remember are for, map, filter/remove
14:44	bhauman	sm0ke: in that tutorial index.html is in the root of your project and so is the out/ directory
14:44	honza	amalloy: cool
14:45	honza	,(chunked-seq? (seq [1 2 3]))
14:45	clojurebot	true
15:14	mpenet	anyone knows if its possible to hide a protocol from codox generated doc?
15:17	mpenet	nm
15:26	yedi	so i have a git repo that has my webapp in it, i gitignore resources/public/out.
15:27	yedi	to deploy my app, I pull the git repo and run lein uberjar from the remote server
15:28	yedi	however when i run lein uberjar, it seems that my resources/public/out isn't getting generated. so i get errors like GET http://reasonedrhymer.com/out/goog/base.js 404 (Not Found)
15:28	dbasch	yedi: it’s usually not a good idea to build things in a production server anyway
15:29	yedi	hm, so I guess I should just somehow get the uberjar to the server during the deploy process? would it make sense to put the uberjar in source control?
15:29	dbasch	yedi: what I do is I have deploy scripts that scp the uberjar to the server
15:30	dbasch	yedi: btw, have you checked the uberjar to make sure the resources are in it?
15:30	yedi	well it ran fine on my vm when I scped the uberjar over for testing
15:31	yedi	is there a way to inspect uberjars for specific resources?
15:31	dbasch	yedi: jar tvf file.jar \|grep resources
15:31	yedi	and yea, looks like i'll just do the scp strat, i was tryna avoid that since i'm generally on shitty internet so it might be slow, but if that's the best way i'll do it
15:31	yedi	cool thanks dbasch
15:53	mpenet	if you use emacs you can just open/browse/modify it with dired
17:33	coventry	Anyone here used double-check? I'm getting "WARNING: No such namespace: math.seedrandom at line 1 file:.../pprng-0.0.2.jar!/cemerick/pprng.cljs" from cljsbuild, then "Undefined nameToPath for math.seedrandom" from javascript. Where should I be pulling in math.seedrandom from? Is it this? https://github.com/davidbau/seedrandom
17:33	cemerick	coventry: follow the directions in the readme re: :libs
18:07	coventry	cemerick: Thanks.
18:11	eraserhd_	So, I've just written some more threading macros. This is usually a bad idea, I imagine.
18:12	eraserhd_	First, I wrote (in-> v a-fn & forms)
18:14	eraserhd_	where, if a-fn is a keyword, is equivalent to (update-in v [a-fn] (fn [s] (-> & forms))).
18:14	eraserhd_	and if a-fn is a function, is equivalent to (a-fn v (fn [s] (-> & forms)))
18:14	eraserhd_	for "lifting" sub-state.
18:16	eraserhd_	Then I wrote ->' , which is like -> except that it treats if and if-not specially, threading the value through the then and else clauses.
18:41	atlasblade	https://www.youtube.com/watch?v=bcuqxj5YTA8
18:41	atlasblade	https://www.youtube.com/watch?v=bcuqxj5YTA8
18:53	ToBeReplaced	is there a function to do (zero 1.0) => 0.0, (zero 1) => 0, etc.?
18:54	Bronsa	ToBeReplaced: (defn zero [n] (- n n))
18:57	coventry	With cemerick.austin, what am I likely doing wrong if the browser repl connection string gives the error 'Cannot write "http://localhost:8001/out/goog/useragent/useragent.js" after document load'?
19:00	dbasch	coventry: what’s your version of clojurescript?
19:00	coventry	eraserhd_: You may want to take a look at synthread. It is a library of threading macros which the authors motivated in terms of monads.
19:00	coventry	Never used it myself.
19:08	tempeh	Hi
19:08	tempeh	I'm looking for a good book or resource on functionally oriented algorithms/patterns
19:08	tempeh	Any ideas?
19:32	waynr	howdy folks
19:33	waynr	anyone have advice for figuring out why this clojure program i am working with takes so long to load?
19:33	nightfly	Is it much longer than any other Java/Clojure program?
19:33	waynr	it's generally around 5 seconds
19:34	nightfly	that sounds pretty normal
19:34	waynr	ah wait i just measured it's more like 15 seconds
19:37	waynr	i added an aleph tcp port to the program which added 10 seconds to the load time compared to what it was previously
19:39	waynr	i mean, an aleph tcp server
19:49	amalloy	waynr: aleph is pretty big. it takes a while to compile all that stuff
19:52	mr-foobar	hi all, I am trying to integrate grunt and cljs https://github.com/harsha-mudi/grunt-contrib-cljs I am getting a "goog not defined" when I run a compiled main.js script.
19:53	mr-foobar	This what I am executing in the grunt `cljsc app/cljs '{:target :nodejs :optimization :none :output-to "generated/js/deps.js" :output-dir "generated/js/"}'`
19:58	dbasch	mr-foobar: how do you run your script?
19:59	mr-foobar	dbasch: `node generated/js/main,js` https://github.com/clojure/clojurescript/wiki/Quick-Start (hello world code)
20:01	dbasch	see “Notes on optimization settings"
20:04	coventry	With test.check, I can I make a generator like gen/any, which goes at least n levels deep before it starts including simple types? I need some way for sized-container to know how deep in the structure it's being called.
20:05	mr-foobar	dbasch: require("./generated/js/goog/bootstrap/nodejs") fails because there is not bootstrap folder
20:06	satshabad	Hi, I've run into a frusterating problem with lein ring server, when I have a dir under /resources/public with a few web fonts ( .eot .svg, .tff and .woff files) lein ring server will auto refresh my browser forever
20:08	satshabad	hmm looks like there is a stylesheet.css in there and when I rm it, the problem stops...
20:10	mr-foobar	dbasch: oops typo, :optimizations :none gives me the bootstrap folder
20:20	owl-v-	assume that the function call is very expensive, how one can remove (recur) tail-recursion?
20:21	tempeh	that seems entirely context dependent
20:22	owl-v-	breath-first-search and depth-first-search?
20:23	owl-v-	and iterating through list and update values?
20:29	owl-v-	ok ok, if i increment a value through loop and append it to a list/vector, this is going to eat up my memory faster than mutable-list/vector.
20:32	owl-v-	if there is a way to remove old values(throw them into garbage collector) then I can happily use immutable list/vector
20:39	crocket	Is clojure going to win on the server side?
20:41	beamso	why not?
20:42	mr-foobar	dbasch: thanks ! the grunt task is working now with :simple need to polish it further.
20:43	tempeh	@owl-v- not sure what you mean by your latter comments, but why do you want to get rid of recur?
20:43	tempeh	Can anyone recommend a good book/resource on functional algorithms/design patterns?
20:44	amalloy	tempeh: well, SICP is free and is amazing. Purely Functional Data Structures has some seriously good stuff but is pretty challenging
20:45	tempeh	@amalloy I've looked into SICP and was turned off by it's use of Sceme. Is it not a problem?
20:45	tempeh	*scheme
20:45	amalloy	language doesn't matter so much. you asked about algorithms and design
20:46	amalloy	and like, it's hard to find a language closer to clojure than scheme anyway
20:46	tempeh	Right, I meant is it a problem if I've never used Scheme
20:46	amalloy	no
20:46	tempeh	did you read it without knowing scheme?
20:46	amalloy	not for long!
20:46	amalloy	pretty soon i knew scheme
20:46	tempeh	ha, okay
20:47	tempeh	thanks, i'll look into getting it then
20:47	amalloy	you can learn scheme in like...an hour, if you're familiar with programming and with lisp especially. the language is so small
20:47	tempeh	in what way is purely functional data structures challenging?
20:47	tempeh	okay
20:48	amalloy	it's a very dense book covering difficult concepts in a way that feels like a mathematics textbook
20:48	tempeh	that might be more what I'm looking for actually
20:48	amalloy	tempeh: you can find some free papers by okasaki and read those, if you want to evaluate whether you want to buy PFDS
20:49	amalloy	actually maybe PFDS is free too? not sure
20:49	amalloy	http://www.cs.cmu.edu/~rwh/theses/okasaki.pdf looks like his thesis - i don't know if the book is identical or just closely related
20:50	agarman	the thesis is more about his framework for analysis of persistent datastructures
20:50	tempeh	ah, okay
20:50	agarman	the book is more about implementations and practical usage
20:51	agarman	both are aws=esmom
20:51	tempeh	I found a directory of his papers, i'll check out one of those
20:51	tempeh	thanks for you help guys, i know what i'm doing tonight
20:51	amalloy	http://www.westpoint.edu/eecs/SiteAssets/SitePages/Faculty%20Publication%20Documents/Okasaki/jfp98sixth.pdf is probably the first thing by him i read
20:51	amalloy	nice and short
20:52	tempeh	oh christ
20:52	tempeh	i hope the answer to "why would anyone ever want to use a 6th order function" is "no one, ever"
20:52	agarman	his thesis is the first thing of his I read
20:53	agarman	flattening combinators is good
20:54	agarman	http://www.westpoint.edu/eecs/SiteAssets/SitePages/Faculty%20Publication%20Documents/Okasaki/jfp03flat.pdf
20:55	amalloy	i should add his disclaimer as my email signature: The views expressed in this [email] are those of the author and do not reﬂect the oﬃcial policy or position of the United States Military Academy, the Department of the Army, the Department of Defense, or the U.S. Government.
20:56	TimMc	ooo, ligature
20:56	TimMc	amalloy: Heck, use that as your IRC signature.
20:57	amalloy	TimMc: maybe my AIM buddy profile
20:57	TimMc	You never know when some statement of yours might be misunderstood as an oﬃcial position of the US gov't.
21:05	owl-v-	tempeh: i want to have some optimization
21:20	owl-v-	it's interesting to see creating a function in a loop in a function have different reference for each iteration: https://www.refheap.com/77755
21:24	xcthulhu	How do I lazily read lines from a gzipped file?
21:29	dbasch	xcthulhu: use line-seq on a java gzipinputstream
21:30	xcthulhu	Trying that, running out of heap :(
21:30	xcthulhu	(defn populate-from-json-gz
21:30	xcthulhu	"Populate the database using a GZIPPED JSON file of chemical reactions"
21:30	xcthulhu	[json-gz]
21:30	xcthulhu	(with-open [data (io/reader
21:30	xcthulhu	(java.util.zip.GZIPInputStream.
21:30	xcthulhu	(io/input-stream json-gz)))]
21:30	xcthulhu	(first (line-seq data))))
21:30	xcthulhu	(populate-from-json-gz “really_big.json.gz")
21:30	xcthulhu	hehe, should have killed the docstring
21:31	dbasch	xcthulhu: use refheap
21:32	xcthulhu	https://www.refheap.com/77756
21:32	xcthulhu	Checking now to make sure that I’m actually writing seperate lines...
21:33	xcthulhu	Since if I don’t have newlines in that huge JSON it would make sense that this is falling over
21:43	xcthulhu	Okay, that was it. Sorry for bothering everyone :(
22:23	Frozenlock	SVG showing as a broken image with compojure. Any ideas why? (All the other types are fine... jpg, png...)
22:24	jdkealy	how am i supposed to get payload params in a post request with compojure ?
22:25	Frozenlock	jdkealy: are you using the params middleware?
22:25	Frozenlock	https://weavejester.github.io/compojure/compojure.handler.html
22:26	jdkealy	i am requiring compojure.handler
22:27	agarman	you need to wrap your routes in wrap-params
22:27	agarman	or api or site
22:27	Frozenlock	But are you using it? (-> my-awesome-app-routes <and now the middleware --->> handler/site)
22:29	jdkealy	no i don't think i am
22:30	jdkealy	i will try to add the middleware to my awesome app routes :)
22:34	dbasch	Frozenlock: how are you serving the file?
22:35	Frozenlock	[:img {:src "/img/some-image.svg"}]
22:35	dbasch	Frozenlock: ok, how are you serving /img
22:36	jdkealy	when you use wrap-params, can you still start server with "lein ring server" or do you need to write a custom start/stop method ?
22:36	Frozenlock	dbasch: You mean like text/html? I let compojure deal with it :p
22:36	dbasch	Frozenlock: I mean, are you using route/resources ?
22:37	Frozenlock	Ah! I'm using the wrap-resource middleware
22:37	Frozenlock	Like I said, everything else seems to work.
22:38	Frozenlock	JS, CSS, images (jpg png)... but not svg
22:38	owl-v-	how do i call java function in myjavacode.java?
22:39	dbasch	Frozenlock: svg is being served as “"image/svg+xml"
22:39	beamso	(package.Class/staticMethod args)
22:39	dbasch	https://github.com/mmcgrana/ring/blob/master/ring-core/src/ring/util/mime_type.clj
22:39	beamso	(instanceMethod instance args)
22:39	Frozenlock	dbasch: Isn't it what it's supposed to be?
22:40	dbasch	Frozenlock: should be, you may want to check
22:41	Frozenlock	uh, my browser says it's being sent with text/xml
22:42	dbasch	Frozenlock: there’s your problem
22:42	dbasch	Frozenlock: try using route/resources instead of wrap-resource
22:47	Frozenlock	dbasch: I would expect them to be quite similar... But will try anyway. Thanks :-)
22:47	dbasch	Frozenlock: I was just looking at the source, route/resources tries to guess the mime type
22:48	dbasch	Frozenlock: I coudn’t find the same thing in wrap-resource
22:48	dbasch	Frozenlock: if you request other file types, what mime-type do you get?
22:48	dbasch	Frozenlock: I’m curious if you get text but the browser recognizes the extension and renders it anyway
22:49	dbasch	“be liberal in what you accept"
22:51	Frozenlock	Hmm... unless my browser is lying, there's image/png and image/jpeg. Local JS is text/plain however...
22:51	Frozenlock	Weird
23:02	Frozenlock	Yay, now my javascript is send with text/javascript.
23:02	Frozenlock	But my svg is still with text/xml
23:06	dbasch	Frozenlock: btw, what browser are you using?
23:06	Frozenlock	chromium
23:08	dbasch	Frozenlock: are you sure the file is proper svg? Chromium renders svg just fine for me even without a mime-type in the response
23:09	dbasch	Frozenlock: what does your main handler look like?
23:09	Frozenlock	Yes, taken directly from wikipedia.
23:10	Frozenlock	https://www.refheap.com/77766
23:10	Frozenlock	I think wrap-content-type middleware isn't added anywhere... O_o
23:10	dbasch	the wikipedia file doesn’t render for me either
23:11	Frozenlock	dbasch: You get improper SVGs from wikipedia?
23:11	Frozenlock	But anyhow, I can see it normaly, but not when I use it on my app.
23:11	dbasch	no, it’s fine
23:11	dbasch	but the wikipedia server serves it with Content-Type:image/svg+xml
23:13	Frozenlock	I'm an idiot
23:13	Frozenlock	resources fixed it alright
23:13	dbasch	If I add (route/resources "/") to my routes, it works
23:13	Frozenlock	It was cached...
23:13	dbasch	and it’s for that reason: wrap-resource does’t guess the mime type, and route-resources does
23:13	dbasch	doesn't
23:14	Frozenlock	F5 instead of shift-F5
23:14	Frozenlock	dbasch: You successfully found and solved my problem, thank you very much! :-)
23:15	dbasch	Frozenlock: np, I wonder if that’s a bug in wrap-resource or the intended behavior
23:15	Frozenlock	Any point in keeping 'wrap-resource' if just 'resources' works?
23:17	dbasch	Frozenlock: resources only works for a specific path, wrap-resource falls back on a handler if no resource is found
23:20	Frozenlock	If it can't find the file, how would falling back on the handler help?
23:21	amalloy	Frozenlock: the point is you use it like (wrap-resource (handle-non-resources))
23:22	Frozenlock	amalloy: Like generating a file on the fly?
23:22	amalloy	so first it tries finding a resource corresponding to the request, and if no such resource exists then it assumes the request must be for something else (eg, one of your dynamically generated pages) and calls your "main" handler
23:22	Frozenlock	I see. Thanks
23:23	amalloy	(wrap-resources foo handler) is a lot like (routes (resources foo) handler)
23:23	dbasch	amalloy: do you think not attempting to guess the mime-type is just an omission?
23:24	amalloy	shrug
23:24	dbasch	amalloy: probably, I see no reason to do it in resources and not in wrap-resource
23:25	amalloy	dbasch: i'm pretty sure weavejester has thought about it a lot more than you or i have, and hundreds of people before you have used this function. if it behaves a particular way, there's probably a reason that's deeper than "i see no reason not to"
23:25	amalloy	but i don't have any special knowledge there, so who knows
23:30	dbasch	amalloy: resources is in compojure, wrap-resource is from ring
23:30	dbasch	there was this issue: https://github.com/ring-clojure/ring/issues/21
23:31	dbasch	but no mention of mime-type
23:31	amalloy	then it's not an omission at all! it's just a feature that's outside of ring's scope, probably
23:31	dbasch	amalloy: makes sense
23:32	Frozenlock	So 'resources' is like wrap-resources, but with a little more sugar
23:36	dbasch	and content-type is too specific for ring, there’s no reason to assume a resource is going to be rendered by a browser at that level