2015-04-04
| 02:46 | justin_smith | tolstoy: you don't need to .stop the server if you pass it var-quoted (run-server #'my-routes {:port 8001}) |
| 02:46 | justin_smith | that way if you redefine my-routes, the new def is used |
| 03:12 | tsdh | How do I specify a java.util.Comparator in Clojure which can be refered to with clojure.api from the Java side (and cast there to Comparator<Object>)? |
| 03:13 | tsdh | I have already a function with args [a b] which returns 0/pos/neg as required for comparators. |
| 03:13 | justin_smith | tsdh: reify java.util.Comparator |
| 03:13 | tsdh | justin_smith: Ah, ok. |
| 03:14 | justin_smith | and use the code for that function, or the function itself, in the compare method |
| 03:15 | amalloy | justin_smith: functions implement Comparator |
| 03:15 | justin_smith | amalloy: oh, wow, I had no idea |
| 03:15 | justin_smith | even better :) |
| 03:15 | justin_smith | (inc amalloy) |
| 03:15 | lazybot | ⇒ 252 |
| 03:15 | justin_smith | actually, you have probably mentioned this before, but I forgot |
| 03:15 | amalloy | probably |
| 03:19 | TEttinger | ,(def people (doto (ArrayList.) (.add "Mike") (.add "Jim") (.add "Bob"))) |
| 03:19 | clojurebot | #error{:cause "Unable to resolve classname: ArrayList", :via [{:type clojure.lang.Compiler$CompilerException, :message "java.lang.IllegalArgumentException: Unable to resolve classname: ArrayList, compiling:(NO_SOURCE_PATH:0:0)", :at [clojure.lang.Compiler analyzeSeq "Compiler.java" 6732]} {:type java.lang.IllegalArgumentException, :message "Unable to resolve classname: ArrayList", :at [clojure.lan... |
| 03:19 | TEttinger | ,(def people (doto (java.util.ArrayList.) (.add "Mike") (.add "Jim") (.add "Bob"))) |
| 03:19 | clojurebot | #'sandbox/people |
| 03:20 | amalloy | TEttinger: (ArrayList. ["Mike" "jim" "Bob"]) |
| 03:20 | TEttinger | ,(Collections/sort people #(> (.length %1) (.length %2))) |
| 03:20 | clojurebot | #error{:cause "No such namespace: Collections", :via [{:type clojure.lang.Compiler$CompilerException, :message "java.lang.RuntimeException: No such namespace: Collections, compiling:(NO_SOURCE_PATH:0:0)", :at [clojure.lang.Compiler analyze "Compiler.java" 6535]} {:type java.lang.RuntimeException, :message "No such namespace: Collections", :at [clojure.lang.Util runtimeException "Util.java" 221]}],... |
| 03:20 | TEttinger | ,(java.util.Collections/sort people #(> (.length %1) (.length %2))) |
| 03:20 | clojurebot | nil |
| 03:21 | TEttinger | ,(vec people) |
| 03:21 | clojurebot | ["Mike" "Jim" "Bob"] |
| 03:21 | amalloy | TEttinger: that doesn't look like a correct comparator |
| 03:21 | justin_smith | it should return an integer, right? |
| 03:21 | amalloy | yes |
| 03:21 | justin_smith | neg, zero, or pos |
| 03:21 | TEttinger | yep, I have never used comparators :| |
| 03:21 | amalloy | justin_smith: though of course ##(doc comparator) exists |
| 03:21 | lazybot | ⇒ "([pred]); Returns an implementation of java.util.Comparator based upon pred." |
| 03:22 | amalloy | to convert a predicate to a comparator |
| 03:22 | tsdh | amalloy, justin_smith: No matter if with plain function or reified comparator, I get "clojure.lang.Var cannot be cast to java.util.Comparator" on the java side. |
| 03:22 | justin_smith | amalloy: wow, cool |
| 03:22 | amalloy | well don't pass it a var |
| 03:22 | tsdh | amalloy: ? |
| 03:22 | justin_smith | tsdh: get the thing the var holds - automatic deref of vars doesn't happen in java |
| 03:22 | TEttinger | ,(java.util.Collections/sort people #(- (.length %1) (.length %2))) |
| 03:22 | clojurebot | nil |
| 03:22 | TEttinger | ,(vec people) |
| 03:22 | clojurebot | ["Jim" "Bob" "Mike"] |
| 03:22 | TEttinger | there we go! |
| 03:22 | justin_smith | tsdh: deref will give you the thing in a var |
| 03:23 | TEttinger | you need an IFn I think |
| 03:23 | tsdh | But there's no deref method (or Var) in clojure.api. All you get is IFn which you can just call() or invoke(). |
| 03:23 | justin_smith | tsdh: uhh... |
| 03:23 | justin_smith | ,(deferf #'+) |
| 03:23 | clojurebot | #error{:cause "Unable to resolve symbol: deferf in this context", :via [{:type clojure.lang.Compiler$CompilerException, :message "java.lang.RuntimeException: Unable to resolve symbol: deferf in this context, compiling:(NO_SOURCE_PATH:0:0)", :at [clojure.lang.Compiler analyze "Compiler.java" 6535]} {:type java.lang.RuntimeException, :message "Unable to resolve symbol: deferf in this context", :at [... |
| 03:23 | justin_smith | err |
| 03:23 | justin_smith | haha |
| 03:24 | justin_smith | ,(deref #'+) |
| 03:24 | clojurebot | #object[clojure.core$_PLUS_ "clojure.core$_PLUS_@13b78423"] |
| 03:24 | TEttinger | was that a definition or a deref, justin_smith? |
| 03:24 | justin_smith | yes, you can deref vars, they are a mutable container |
| 03:24 | amalloy | the clojure.java.api.Clojure API is very minimal |
| 03:25 | justin_smith | ,(.deref #'+) |
| 03:25 | clojurebot | #object[clojure.core$_PLUS_ "clojure.core$_PLUS_@13b78423"] |
| 03:25 | justin_smith | you can use the deref method too |
| 03:25 | amalloy | if you have a var which contains a value, and you need the value instead of the var (because it's not a function, you'll have to get the var for deref, and call it on your var |
| 03:25 | amalloy | justin_smith: c.j.api.Clojure gives you everything as IFn, not even Var |
| 03:25 | justin_smith | ahh |
| 03:25 | amalloy | so you'd have to cast it to var and then call .deref |
| 03:26 | amalloy | but better to find the var for deref and use that |
| 03:26 | tsdh | amalloy: Got the same error when my clojure-defined comparator as a plain function. |
| 03:26 | justin_smith | interesting |
| 03:27 | tsdh | Hm, indeed (supers (class (fn [a b]))) says it implements Comparator. |
| 03:28 | justin_smith | tsdh: and you are definitely getting the fn and not the var holding it? |
| 03:28 | tsdh | I do final static IFn MATCH_COMPARATOR = Clojure.var(SOLUTION_NS, "match-comparator"); in java. |
| 03:28 | justin_smith | yeah, that's the var (as IFN, which is a super of var) |
| 03:29 | tsdh | I did the same with other fns which I can myIFn.invoke(args); in java. |
| 03:29 | justin_smith | tsdh: yeah, calling a var implictly derefs when invoking |
| 03:29 | justin_smith | ,(#'+ 1 2) |
| 03:29 | tsdh | I guess I could invoke my comparator with myComp.invoke(a, b); but I need to pass it on to some framework. |
| 03:29 | clojurebot | 3 |
| 03:31 | amalloy | tsdh: (Comparator)deref.invoke(myvar) |
| 03:31 | tsdh | Aha: comparator = (Comparator<Object>) ((Var) FunnyQTBenchmarkLogic.MATCH_COMPARATOR).deref(); |
| 03:32 | tsdh | amalloy: Ah, requirirng deref and invoking it on my var would also work. Anyway, gotta run. Thank you both! |
| 03:32 | amalloy | well i did suggest that a few times |
| 03:33 | justin_smith | haha |
| 03:33 | justin_smith | that amalloy, he gets no respect I tells ya what |
| 03:34 | amalloy | all i get is this useless karma |
| 03:35 | TEttinger | (dec amalloy's respect) |
| 03:35 | lazybot | ⇒ -1 |
| 03:35 | TEttinger | (inc amalloy) |
| 03:35 | lazybot | ⇒ 253 |
| 03:35 | TEttinger | less than no respect! |
| 05:31 | constl | What does the "&" mean in a general form? e.g. (while test & body) |
| 05:32 | TEttinger | constl, it means that you can pass additional args after the & and they will be available as a seq with the name body (there, at least) |
| 05:32 | TEttinger | so "& args" in the typical clojure -main fn means you can pass a number of arguments and get them from a seq called args |
| 05:33 | constl | TEttinger: So in my example it means that body can have multiple s-expressions ? |
| 05:34 | TEttinger | yes, since an s-expression is just a form. hm, that isn't a an argvector though so.... |
| 05:34 | TEttinger | actually I haven't seen that in a general form |
| 05:34 | TEttinger | can you link to the code where that was? |
| 05:35 | TEttinger | it's possible there's some surrounding macro that defines & in its body |
| 05:35 | TEttinger | (doc while) |
| 05:35 | clojurebot | "([test & body]); Repeatedly executes body while test expression is true. Presumes some side-effect will cause test to become false/nil. Returns nil" |
| 05:35 | constl | This reference is from a book called "Clojure in action" and it is explaining the functional iteration capabilities |
| 05:35 | constl | so to examplain the while general form it gives that example |
| 05:37 | constl | Sorry my English doesn't make any sense in the above two sentences. But i think you've answered my question already |
| 05:37 | TEttinger | ,(while (not= 0 (rand-int 10)) (def inner-form 1) (def another 2) (+ inner-form another)) |
| 05:37 | clojurebot | nil |
| 05:37 | TEttinger | ,(while (not= 0 (rand-int 10)) (def inner-form 1) (def another 2) (print (+ inner-form another))) |
| 05:37 | clojurebot | 333333333 |
| 05:38 | TEttinger | I think the & is a typo in there |
| 05:38 | TEttinger | it makes sense in an argument list |
| 05:38 | TEttinger | well, argument vector |
| 05:38 | TEttinger | unless it's using that as pseudocode to mean one or more forms in body |
| 05:39 | constl | Well it is also found when describing the "when". e.g. (when test & body) |
| 05:39 | TEttinger | (which is totally useful to be able to write that, but they should have explained it better) |
| 05:39 | TEttinger | yeah, when also allows multiple forms, unlike if |
| 05:39 | constl | and it gives this explanation, (when true (do-this) (and-that) (and-return-this)) |
| 05:39 | TEttinger | yep |
| 05:40 | TEttinger | ,(if true 15 :not-found :something-not-allowed) |
| 05:40 | clojurebot | #error{:cause "Too many arguments to if", :via [{:type clojure.lang.Compiler$CompilerException, :message "java.lang.RuntimeException: Too many arguments to if, compiling:(NO_SOURCE_PATH:0:0)", :at [clojure.lang.Compiler analyzeSeq "Compiler.java" 6732]} {:type java.lang.RuntimeException, :message "Too many arguments to if", :at [clojure.lang.Util runtimeException "Util.java" 221]}], :trace [[cloju... |
| 05:40 | TEttinger | ,(when true 15 :not-found :something-that-is-allowed) |
| 05:40 | clojurebot | :something-that-is-allowed |
| 05:41 | TEttinger | but when is pretty much exactly the same as (if test (do & body)) |
| 05:42 | TEttinger | the only difference between an if with a do block for the "test is true" condition and no "test is false" block... |
| 05:42 | constl | ,(if true (do 15 :not-found) :something-that-is-allowed) |
| 05:42 | clojurebot | :not-found |
| 05:42 | TEttinger | the only one i can think of is that when is usually used by clojure programmers to show that it's being used for something stateful or side-efecting |
| 05:43 | TEttinger | exactly |
| 05:43 | TEttinger | and do is the simplest example of what the book would use & for |
| 05:43 | TEttinger | (do & body) |
| 05:43 | TEttinger | it just does every form in the body |
| 05:43 | TEttinger | in order |
| 05:43 | constl | Right! |
| 05:44 | constl | Cheers for the insight |
| 05:47 | TEttinger | great! clojure's a fun language. (I think the most fun part is when someone else is writing java code and your clojure version ends up being a quarter the size) |
| 05:48 | TEttinger | (not accurate in the general case, I've had it be a fifth sometimes and 2/3 maybe sometimes) |
| 05:48 | constl | TEttinger: Well coming from a Java background and being an average Joe i would say it boils my brain |
| 05:52 | TEttinger | it was hard for me for a long time, tbh. it did click, in parts, the more I wrote |
| 05:52 | TEttinger | 4clojure helped, if I was starting today I'd read the heck out of brave clojure |
| 05:57 | constl | TEttinger: Well i started with brave clojure but somehow i felt it moved too fast so i resorted to Clojure in action and it seems its ok for me. |
| 05:57 | constl | What's 4clojure ? |
| 05:57 | TEttinger | it's an excellent koan-like site. it presents you with fill-in-the-blank problems |
| 05:58 | TEttinger | you write code that solves all of the blanks on the page |
| 05:58 | TEttinger | like one could be simple, (= 2 (+ _ 1)) |
| 05:58 | TEttinger | so for that one you just enter 1 |
| 05:59 | constl | I ll check that then, sounds fun |
| 05:59 | TEttinger | the next will be trickier, like (= "Hello, World!" (_ "World")) |
| 06:00 | TEttinger | where you would need to fill in a function that it would run with the arg "World" |
| 06:01 | TEttinger | by maybe 10-20 problems in (maybe earlier, I don't remember) they start being very tricky and challenging. a lot have you re-implement fns from clojure's library, without using those fns or sometimes certain similar ones |
| 06:01 | TEttinger | the one that has you reimplement nth is a fun one |
| 06:02 | constl | Do i have to register to start ? |
| 06:03 | constl | I mean when you registered you are presented with a sequence of problems or is it just that you simply select which one to solve? |
| 06:03 | TEttinger | yeah, it's a good idea to register to keep track of your progress |
| 06:03 | TEttinger | you usually start with Number 1 |
| 06:03 | TEttinger | (not sure if there's a 0) |
| 06:04 | TEttinger | a good piece of info that may be handy for the one with nth: several data structures clojure provides are also usable as functions, like maps can be used as functions that take a key as an argument and try to get that key in the map |
| 06:04 | TEttinger | ,({"alpha" 1 "beta" 2} "alpha") |
| 06:04 | clojurebot | 1 |
| 06:05 | TEttinger | experiment a bit with what you can do there, trying a list, a vector, and a set for example |
| 06:07 | constl | but it will need to go hand by hand with some learning material, it's a learning source by itself right? |
| 06:08 | TEttinger | yeah, you would still need the book |
| 06:08 | TEttinger | it's more practice or exercises to test your knowledge in ways that you might not get otherwise |
| 06:09 | constl | hmm |
| 06:10 | constl | is there a way to submit its form without the mouse ? like a keyboard shortcut ? |
| 06:11 | TEttinger | might be a good issue report, Raynes I think is a primary author |
| 08:25 | constl_ | ,(println "Hello chan") |
| 08:25 | clojurebot | Hello chan\n |
| 08:26 | constl_ | Im trying to create a simple function that you can pass two numbers and it will countdown from first to second |
| 08:26 | gfredericks | countdown like by printing? |
| 08:27 | constl_ | gfredericks: Yes, and i'm using loop and recur but somehow the code i came up with doesn't look very optimal |
| 08:27 | constl_ | Can i paste here for your opinion to this trivial question ? |
| 08:27 | gfredericks | use refheap.com |
| 08:29 | constl_ | https://www.refheap.com/4f21f193624bf80669e7e6c16 |
| 08:30 | gfredericks | constl_: the loop part is redundant, you can recur to the function itself |
| 08:31 | gfredericks | you can also replace (if c (do ...)) with (when c ...) |
| 08:31 | gfredericks | not sure what else you don't like abou it |
| 08:32 | constl_ | gfredericks: So if i understood correctly its better to remove loop, and rebind decremented value to x and call that function again ? |
| 08:35 | constl_ | Something like this https://www.refheap.com/99226 |
| 09:03 | gfredericks | constl_: yeah it's a little simpler |
| 09:03 | gfredericks | constl_: you might also look at range and doseq for an alternative approach |
| 10:11 | catern | dear clojure fans |
| 10:11 | catern | how would I idiomatically get the sequence of partial sum of another sequence? |
| 10:12 | catern | i.e.: [1 2 3 4 5] -> [1 3 6 10 15] |
| 10:12 | catern | (preferably lazily) |
| 10:26 | dysfun | catern: i'd probably do something like this: |
| 10:26 | dysfun | ,(->> [1 2 3] (reductions conj []) rest (map (partial apply +))) |
| 10:26 | clojurebot | (1 3 6) |
| 10:26 | catern | :( |
| 10:26 | catern | that does a lot of unnecessary work |
| 10:27 | catern | wait! |
| 10:27 | catern | dysfun: you're being silly! |
| 10:27 | catern | reductions is the function to use |
| 10:27 | catern | that's all I was looking for :) |
| 10:27 | catern | ,(reductions + [1 2 3 4 5]) |
| 10:27 | clojurebot | (1 3 6 10 15) |
| 10:28 | dysfun | aha, that's pretty nice |
| 10:28 | dysfun | i haven't woken up properly today :) |
| 10:28 | dysfun | i'm in grunt mode, so i'm writing tests |
| 10:35 | justin_smith | ,(reductions + [1 2 3 4 5]) |
| 10:35 | clojurebot | (1 3 6 10 15) |
| 10:35 | justin_smith | ^ catern |
| 10:35 | justin_smith | oh, never mind |
| 10:35 | justin_smith | haha |
| 10:37 | expez | I think I'm making this harder than it should be, how can I get the content of a clj file without the (ns ..) form? |
| 10:38 | justin_smith | expez: as a string? |
| 10:38 | expez | justin_smith: yes |
| 10:41 | expez | My first attempt skips into the ns form and then drops tokens from the string until a paren count of 0 is reached. But then I realized parens might occur (without being balanced) in the docstring so I would have to account for that. |
| 10:41 | catern | on the off chance |
| 10:42 | catern | does anyone/can quickly think of a way to nicely and idiomatically do a queueing process simulation? |
| 10:42 | justin_smith | yeah, it's not an easy problem |
| 10:42 | catern | brb, /me goes off to think about it |
| 10:42 | justin_smith | sorry, expez: that is not an easy problem |
| 10:43 | justin_smith | catern: queueing process simulation, I would use either agents (which are all backed by queues they use to implement send / send-off) or core.async (channels are queues) |
| 10:44 | catern | justin_smith: not quite like that :) |
| 10:44 | gfredericks | expez: read can do this actually |
| 10:44 | justin_smith | oh, of course |
| 10:44 | catern | justin_smith: as in, I draw values from a distribution and accumulate them to get arrival times |
| 10:44 | catern | but I just realized that I could search for examples in Haskell or something |
| 10:44 | gfredericks | ,(def my-file (java.io.StringReader. "(ns foo) (hey) (hooray)")) |
| 10:44 | clojurebot | #'sandbox/my-file |
| 10:44 | gfredericks | ,(read my-file) |
| 10:44 | clojurebot | #error{:cause "java.io.StringReader cannot be cast to java.io.PushbackReader", :via [{:type java.lang.ClassCastException, :message "java.io.StringReader cannot be cast to java.io.PushbackReader", :at [clojure.core$read invoke "core.clj" 3630]}], :trace [[clojure.core$read invoke "core.clj" 3630] [clojure.core$read invoke "core.clj" 3628] [clojure.core$read invoke "core.clj" 3626] [sandbox$eval47 i... |
| 10:45 | gfredericks | ,(def my-file (java.io.PushbackReader. (java.io.StringReader. "(ns foo) (hey) (hooray)"))) |
| 10:45 | clojurebot | #'sandbox/my-file |
| 10:45 | gfredericks | ,(read my-file) |
| 10:45 | clojurebot | (ns foo) |
| 10:45 | gfredericks | ,(read my-file) |
| 10:45 | clojurebot | (hey) |
| 10:45 | gfredericks | ,(read my-file) |
| 10:45 | clojurebot | (hooray) |
| 10:45 | gfredericks | expez: ^ |
| 10:45 | expez | gfredericks: that was my first though, to read the entire file and drop the first form I read, but that would change the formatting of the file when I spit it back out. |
| 10:45 | justin_smith | catern: ahh, not processes with queues for communication, but simulating the process of a queue |
| 10:45 | catern | justin_smith: indeed |
| 10:46 | gfredericks | expez: maybe you want a reader that knows how much has been read |
| 10:46 | justin_smith | catern: if you had an ordered list of entities, and some way to calculate how long each takes to get served, I would use reductions to get a series of wait times |
| 10:46 | gfredericks | there's gotta be something for that; all you need to know is the position of that last paren |
| 10:47 | expez | gfredericks: yeah |
| 10:47 | justin_smith | expez: shell out to elisp :P find the matching paren to the ns form, and drop the form |
| 10:47 | justin_smith | or call read to drop the ns form, and then slurp to get the rest unchanged, maybe |
| 10:48 | expez | justin_smith: the current implementation uses mechanical turk for human assistance, but since the op is sync this caused some users to complain about performance |
| 10:48 | gfredericks | ,(def my-file (java.io.PushbackReader. (java.io.StringReader. "(ns foo) (hey) (hooray)"))) |
| 10:48 | clojurebot | #'sandbox/my-file |
| 10:48 | gfredericks | ,(read my-file) |
| 10:48 | clojurebot | (ns foo) |
| 10:48 | gfredericks | (slurp my-file) |
| 10:48 | gfredericks | ,(slurp my-file) |
| 10:48 | clojurebot | " (hey) (hooray)" |
| 10:48 | gfredericks | nice |
| 10:48 | justin_smith | there we bo |
| 10:48 | justin_smith | haha, go |
| 10:49 | expez | damn, that's hot stuff. Thanks! |
| 11:21 | catern | okay, I have an idea for my queue simulation |
| 11:22 | catern | but I need to map over two seqs, and my mapping function needs to know the last value I created |
| 11:22 | catern | ,(map f (range 5) (range 5 10)) |
| 11:22 | clojurebot | #error{:cause "Unable to resolve symbol: f in this context", :via [{:type clojure.lang.Compiler$CompilerException, :message "java.lang.RuntimeException: Unable to resolve symbol: f in this context, compiling:(NO_SOURCE_PATH:0:0)", :at [clojure.lang.Compiler analyze "Compiler.java" 6535]} {:type java.lang.RuntimeException, :message "Unable to resolve symbol: f in this context", :at [clojure.lang.Ut... |
| 11:23 | catern | ,(map + (range 5) (range 5 10)) |
| 11:23 | clojurebot | (5 7 9 11 13) |
| 11:23 | catern | but + also needs to have access to 5 while calculating 7, 7 while calculating 9, etc. |
| 11:23 | catern | should I just mantain some mutable variable or is there a nice way to do that? |
| 11:23 | justin_smith | ,(map list (partition 2 1 (range 5)) (partition 2 1 (range 5 10))) |
| 11:23 | clojurebot | (((0 1) (5 6)) ((1 2) (6 7)) ((2 3) (7 8)) ((3 4) (8 9))) |
| 11:24 | gfredericks | does that sound like reduce at all? |
| 11:24 | justin_smith | gfredericks: sounds like a reduce to me yeah |
| 11:24 | catern | gfredericks: a little, but doing it with reduce would also be weird |
| 11:24 | catern | because I'd need to recalculate the last value every time |
| 11:25 | catern | (which would be fine, it'd just be weird... also what if there was randomness or state in the calculation of the last value?) |
| 11:25 | justin_smith | catern: you can do whatever you want to calculate each value, but the previous is an input, sounds like reductions |
| 11:26 | catern | justin_smith: not sure how reductions is any different from reduce here |
| 11:26 | justin_smith | catern: gives you the series |
| 11:27 | catern | yeah, but I can't access that series while I'm calculating it... well, I could, I suppose, but specifically I can't access the last element of that series |
| 11:27 | justin_smith | ,(reductions (fn [prev [a b]] (+ (/ prev 10.0) a b)) 0 (map list (range 5) (range 5 10))) |
| 11:27 | clojurebot | (0 5.0 7.5 9.75 11.975 ...) |
| 11:27 | justin_smith | catern: reduce makes it easy to access your previous result |
| 11:27 | justin_smith | as does reductions |
| 11:28 | catern | hmm |
| 11:28 | catern | I'll try it |
| 11:28 | justin_smith | catern: and you absolutely can always get the previous value returned, it is the guaranteed first arg to your function in reductions or reduce |
| 11:33 | justin_smith | catern: so you want a running sum but also the previous args? |
| 11:33 | catern | justin_smith: yes, I think |
| 11:34 | justin_smith | ,(reductions (fn [[sum prev] [a b]] [(+ sum a b) [a b]] [0 0] (map list (range 5) (range 5 10))) |
| 11:34 | clojurebot | #<RuntimeException java.lang.RuntimeException: EOF while reading> |
| 11:34 | justin_smith | ,(reductions (fn [[sum prev] [a b]] [(+ sum a b) [a b]]) [0 0] (map list (range 5) (range 5 10))) |
| 11:34 | clojurebot | ([0 0] [5 [0 5]] [12 [1 6]] [21 [2 7]] [32 [3 8]] ...) |
| 11:35 | justin_smith | second one should have had nil as the init |
| 11:35 | justin_smith | but hopefully the idea should be clear - first item in pair is the running sum, second item is the previous pair |
| 11:44 | dysfun | win 20 |
| 11:45 | catern | hey, yeah, reductions works perfectly, thanks justin_smith |
| 11:45 | justin_smith | np |
| 11:48 | wirrbel | Running lein ring server-headless, shouldn't my server just rebuild upon source changes? |
| 11:49 | justin_smith | no, it doesn't and won't |
| 11:50 | justin_smith | wirrbel: if you pass your app/handler whatever as a var, the server will see updates to the definition though |
| 11:50 | justin_smith | eg. #'my-app.core/handler instead of my-app.core/handler |
| 12:24 | gfredericks | when passed no args, alter-var-root returns a destructive transducer that just messes everything up |
| 13:45 | CL4Y3RSCLUB | SO? |
| 13:52 | justin_smith | ~so |
| 13:52 | clojurebot | so is (add-to-list 'erc-keywords '("\\bso\\b" erc-default-face)) |
| 13:52 | Bronsa | makes sense clojurebot |
| 13:53 | justin_smith | oh, of course, that's how to unhighlight so, who isn't around anymore |
| 13:53 | justin_smith | $karma so |
| 13:53 | lazybot | so has karma -34. |
| 13:53 | justin_smith | the poor sould |
| 13:53 | justin_smith | *soul |
| 13:54 | Bronsa | lol |
| 14:08 | catern | (inc clojure) |
| 14:08 | lazybot | ⇒ 21 |
| 14:09 | justin_smith | $karma Clojure |
| 14:09 | lazybot | Clojure has karma 21. |
| 14:09 | justin_smith | ahh, ignores caps |
| 14:09 | justin_smith | (inc clojure) |
| 14:09 | lazybot | ⇒ 22 |
| 14:09 | justin_smith | $karma Clojure |
| 14:09 | lazybot | Clojure has karma 22. |
| 14:14 | justin_smith | noncom: I found support for unquoted field names in cheshire |
| 14:14 | justin_smith | noncom: https://github.com/dakrone/cheshire/blob/master/src/cheshire/factory.clj#L15 |
| 14:30 | gfredericks | justin_smith: oh hey look at those dynamic vars |
| 14:30 | gfredericks | I'm trying to figure out if they can be used to solve the global extensibility problem |
| 14:30 | justin_smith | gfredericks: that's what I was looking for when I found those actually :) |
| 14:31 | gfredericks | ,'com.fasterxml.jackson.core.JsonFactory |
| 14:31 | clojurebot | com.fasterxml.jackson.core.JsonFactory |
| 14:31 | justin_smith | sadly I don't think the coercions are per-factory |
| 14:31 | gfredericks | ,com.fasterxml.jackson.core.JsonFactory |
| 14:31 | clojurebot | #error{:cause "com.fasterxml.jackson.core.JsonFactory", :via [{:type clojure.lang.Compiler$CompilerException, :message "java.lang.ClassNotFoundException: com.fasterxml.jackson.core.JsonFactory, compiling:(NO_SOURCE_PATH:0:0)", :at [clojure.lang.Compiler analyze "Compiler.java" 6535]} {:type java.lang.ClassNotFoundException, :message "com.fasterxml.jackson.core.JsonFactory", :at [java.net.URLClassL... |
| 14:31 | gfredericks | &com.fasterxml.jackson.core.JsonFactory |
| 14:31 | lazybot | ⇒ com.fasterxml.jackson.core.JsonFactory |
| 14:31 | justin_smith | if they were, yeah, the logger could have its own factory |
| 14:31 | gfredericks | &(.isInterface com.fasterxml.jackson.core.JsonFactory) |
| 14:31 | lazybot | ⇒ false |
| 14:32 | gfredericks | hmph |
| 14:33 | gfredericks | yeah I guess if cheshire coerces everything beforehand |
| 14:33 | justin_smith | yeah, it uses the JSONable protocol, with a single method, to-json |
| 14:33 | justin_smith | and that can't be a local binding |
| 14:35 | justin_smith | and clojure.data.json has the JSONWriter protocol, with -write method |
| 14:37 | gfredericks | I feel like I asked stu about this assymetry a couple years back, but have no idea what he said |
| 14:40 | gfredericks | maybe that was more about data readers |
| 14:42 | gfredericks | I guess transit avoids the global thing a bit better than clojure's printer does |
| 14:45 | justin_smith | gfredericks: oh, that's a great point, maybe transit would be the best backend for this |
| 14:45 | gfredericks | what |
| 14:45 | gfredericks | for what |
| 14:45 | gfredericks | what are we |
| 14:45 | gfredericks | is this even talking about |
| 14:48 | justin_smith | never mind, I may have been confusing transit for something else |
| 14:48 | justin_smith | I thought it would work with various formats, but no |
| 14:48 | justin_smith | that's not the one |
| 14:49 | gfredericks | no transit is like a frontend |
| 15:00 | tomjack | isn't transit what happens between ends? |
| 15:01 | tomjack | (and starts and stops I suppose) |
| 15:03 | justin_smith | tomjack: there is a specific lib for sending data between processes called transit |
| 15:03 | tomjack | yes :) |
| 15:04 | tomjack | but the name was chosen with some care, I presume |
| 15:17 | matthavard | How would I add this library as a dependency: https://github.com/dcolthorp/matchure |
| 15:17 | matthavard | oh oops there's an installation section ignore me |
| 15:18 | justin_smith | matthavard: have you tried [matchure "0.10.1"] ? that's what it has in the project.clj |
| 15:18 | justin_smith | oh yeah, that's what it would tell you to do also |
| 15:18 | matthavard | Yeah haha I'm dumb I was trying [org.clojure/matchure "0.10.1"] |
| 17:20 | sveri | Hi, how can I define a return type of Foo[] for a clojure method? (defn ^Foo[] fname seems not to work for me |
| 17:31 | scottj | sveri: maybe ^"[Lnamespace.Foo;" |
| 17:33 | sveri | scottj: I just found some thread from 2009 suggesting this should work: #^"[Lnamespace.Foo;" But your one looks like it works in the latest clojure, at least that how ever return type definitions are declard |
| 17:33 | sveri | Thank you very much |
| 17:35 | scottj | ^ used to be #^, afair |
| 17:35 | sveri | Yea, that makes sense :-) |
| 18:47 | noncom|2 | is there any particular reason on why a lein's project file is a call to some (defproject ..) instead of being a simple edn {} ? |
| 18:49 | tolstoy | noncom|2: Total guess, but defproject is a macro and probably does some lightweight transformation? |
| 18:50 | justin_smith | yeah, I think that's pretty much it |
| 18:50 | noncom|2 | tolstoy: yeah, sure, i think.. but i totally see no reason for it being this way. afaik project.clj is a plain data map, so why complicate the file with a call to some fn or macro? |
| 18:50 | justin_smith | tolstoy: it supports heavy weight transformation via ~ |
| 18:50 | noncom|2 | ah, so it's about ~ |
| 18:50 | justin_smith | that could be part of it, at least |
| 18:50 | justin_smith | you can evaluate arbitrary code in there |
| 18:51 | noncom|2 | uh.. |
| 18:51 | noncom|2 | does not seem to clean though.. but probably useful at times |
| 18:51 | noncom|2 | s/to/too |
| 18:51 | justin_smith | I'm not saying it's a good idea |
| 18:51 | justin_smith | I am saying it's possible |
| 18:52 | justin_smith | and it would make reading a project.clj as edn a problem |
| 18:52 | justin_smith | also, a project.clj is only the parts of the config you altered, it gets merged with the base options |
| 18:52 | justin_smith | unless you used ^:replace, etc. etc. |
| 18:53 | justin_smith | btw, noncom|2, did you see my link earlier about being able to replace cheshire's json-factory with one that supports raw keys in maps? |
| 18:53 | justin_smith | it's in factory.clj |
| 18:54 | noncom|2 | i did not see the link! but i know that it is possible to (binding ..) factory for a json operation, but it only affects reading.. |
| 18:54 | noncom|2 | was there something about writing a json string too? |
| 18:54 | justin_smith | factory is for generating |
| 18:55 | justin_smith | well, maybe not only for generating, but it definitely affects generation |
| 18:55 | tolstoy | You can have multiple forms in project.clj, so one of 'em has to be tagged as the project. |
| 18:55 | noncom|2 | it's like string -> factory -> clojure data struct |
| 18:55 | justin_smith | it's also like clojure data struct -> factory -> string |
| 18:55 | noncom|2 | wow, did not notice that this was possible |
| 18:55 | tolstoy | https://github.com/technomancy/leiningen/blob/master/leiningen-core/src/leiningen/core/project.clj#L398-L405 Suggests that (def project {}) works, too. |
| 18:55 | noncom|2 | do you still have the link? |
| 18:56 | justin_smith | https://github.com/dakrone/cheshire/blob/master/src/cheshire/factory.clj#L15 |
| 18:56 | justin_smith | tolstoy: I am enlightened but not surprised :) cool |
| 18:56 | noncom|2 | tolstoy: yeah, looks like we have the potential to write another OS with lein projects :D |
| 18:56 | justin_smith | haha |
| 18:57 | justin_smith | noncom|2: note that the comments there explicitly state that the factory is used for both encoding and decoding |
| 18:59 | noncom|2 | justin_smith: so, this factory, yes.. i tried it, but i only managed to make cheshire use it when reading.. the properties are mapped to JsonParser class.. which sounds like it parses the string. however, right.. the comment says both enc/dec.. |
| 18:59 | noncom|2 | i must give it a second, closer look at monday |
| 19:00 | noncom|2 | playing with neo4j, yeah, in making a little mudy thing :) |
| 19:00 | noncom|2 | this neo4j takes in this wrong kind of json as parameters for its objects. the json kind without the quotes on fields names |
| 19:06 | underplank | So Im trying to get a repl working in emacs using cider-jack-in command. And the repl seems to be starting, but I can get the buffer to open. Any ideas how to get that to happen? |
| 19:07 | justin_smith | underplank: can you use M-x cider to connect to a repl that is already open? |
| 19:09 | underplank | hmm… thats wierd. I dont get a prompt or anything. |
| 19:09 | underplank | It looks like I have the connections, but nothing executes. |
| 19:12 | justin_smith | underplank: has cider ever worked for you in the past? do you have the cider-nrepl plugin set up? |
| 19:13 | underplank | I’ve had it working in the past. lein deps :tree shows that I have 0.8.2 installed. I do that in my lein profiles.clj |
| 19:13 | underplank | Do I need to have that in my project.clj as well? |
| 19:13 | justin_smith | no |
| 19:14 | underplank | hmm.. I didnt think so. |
| 19:14 | justin_smith | if it has worked in the past, the likely problem is that you upgraded your cider elisp without deleting the elc files |
| 19:14 | justin_smith | cider breaks binary compatibility, and elc doesn't handle that nicely |
| 19:17 | underplank | ahh. that might make sense. I think my emacs.d folder is in a … not great state.. |
| 19:18 | justin_smith | find . -name '*.elc' will show you where they are |
| 19:18 | justin_smith | because of how elisp works, it can be effected by elc files that are not part of cider proper, if they call things in cider |
| 19:19 | underplank | yeah looks like I have like three versions in here or something. |
| 19:19 | underplank | whoops! |
| 19:19 | justin_smith | when I said "binary compatibility" above, I should have said API compatibility (argument types, counts, etc. etc.) |
| 19:19 | justin_smith | well, that can be fine, as long as you are loading the right one |
| 19:19 | justin_smith | heh |
| 19:20 | underplank | are elc files just like compiled version of source, so If I blow them away it will rebuild the ones it needs? |
| 19:20 | justin_smith | they are like aot |
| 19:20 | justin_smith | if they don't exist, emacs will just use the source files |
| 19:20 | justin_smith | and startup may take a little longer |
| 19:20 | underplank | ahh great.. thats what I thought. |
| 19:27 | underplank | justin_smith: works! just cleaned up a few packages and installed the latest.. thanks! |
| 19:27 | justin_smith | np |
| 19:28 | bordatoue | could anyone please tell me how to change nrepl version 0.2.6 to 0.2.8 under lein , currently I had created ~/.lein/profiles.clj file with the following setting {:user {:plugins [[cider/cider-nrepl "0.9.0-SNAPSHOT"]] |
| 19:28 | bordatoue | :dependencies [[org.clojure/tools.nrepl "0.2.7"]]}} |
| 19:29 | justin_smith | why do you ask for 0.2.7 ? |
| 19:29 | bordatoue | ok need to change it from 0.2.6 --> 0.2.7 ; the |
| 19:29 | bordatoue | the problem is it downloads the 0.2.7 jar but still startsup with 0.2.6 |
| 19:30 | tolstoy | bordatoue: It's up to 0.2.10 now. |
| 19:30 | bordatoue | it seems to only work if I add the dependencies under the project.clj file |
| 19:30 | bordatoue | tolstoy: the point is i can't get it to work with any version other than 0.2.6 it keeps chaning every day |
| 19:31 | justin_smith | bordatoue: there's a weird issue, you may also be able to use :dev rather than :user in profiles.clj to get the proper version |
| 19:31 | tolstoy | bordatoue: I added 0.2.10 under {:dev {:dependencies ...}} as it's just used for nrepl. Seems to work okay. |
| 19:32 | bordatoue | i can't understand this why they keep chaning this all the time , everytime i update my emacs then the entire repl breaks |
| 19:32 | bordatoue | tolstoy: I will try that, thanks a lot |
| 19:32 | justin_smith | bordatoue: this is why I stopped using cider, the development is too chaotic |
| 19:32 | justin_smith | tolstoy: yeah, that's what I was trying to suggest :) |
| 19:32 | bordatoue | justin_smith: what do you use |
| 19:33 | justin_smith | bordatoue: inferior-lisp, which has 0 features, but it actually works |
| 19:33 | bordatoue | have you tried intellij plugin for clojure is it any good |
| 19:33 | justin_smith | eg. I don't even get completion or anything |
| 19:33 | bordatoue | i don't mind any feature but consistency is my top priority |
| 19:33 | justin_smith | bordatoue: I'm too much of an emacs addict, but I have heard nothing but good things about it, and have watched people use it and been pretty impressed |
| 19:34 | justin_smith | cursive that is |
| 19:34 | mfikes | bordatoue: Yes, Cursive is great. |
| 19:34 | justin_smith | bordatoue: yeah, I switched to inferior-lisp because I was tired of the break / update cycle |
| 19:34 | bordatoue | great guys, I will give it a try |
| 19:35 | tolstoy | I stuck with 0.8.2 for a long while. Seemed stable to me. |
| 19:35 | tolstoy | inf-clojure has its appeal. |
| 19:38 | bordatoue | tolstoy: can you please paste what you have entered in your profile.clj file |
| 19:38 | bordatoue | i tried :dev :dependencies and it is still not working |
| 19:38 | tolstoy | bordatoue: Sure, but I don't use profiles.clj for that anymore. |
| 19:39 | bordatoue | then what do you use ? are u adding it to the project.clj file |
| 19:39 | tolstoy | What's the refthingy we can paste to? |
| 19:39 | justin_smith | http://refheap.com |
| 19:40 | freddd | can anyone tell me if transducers and reducers play nice? as in, is it possible to use transducers for most of a pipeline, but use reducers/fold for the final combination for the fork/join parallelism benefits? |
| 19:40 | tolstoy | https://www.refheap.com/99240 |
| 19:41 | tolstoy | bordatoue: Here's a more complete :profiles vector in my project.clj file (for a super new project): https://www.refheap.com/99241 |
| 19:42 | underplank | Hi all.. When using uberjar what is the difference between the standalone jar and the uhh non-standalone jar? |
| 19:42 | freddd_ | dang, got cut off just after asking my question regarding reducers/transducers. if anyone replied, would you mind replying again? :) |
| 19:42 | bordatoue | so everytime you create a project you need to explicitly add this |
| 19:43 | justin_smith | underplank: one contains your project, the other contains your project plus all deps, it uses the project one as part of constructing the other one |
| 19:43 | tolstoy | bordatoue: I do that just for my own self so that I'm super clear what's interfering with my project. Personal taste. |
| 19:43 | underplank | ahh.. is there a reason you would want to use the non-standalone? |
| 19:43 | justin_smith | freddd_: as long as you are using transducers and reducers properly, they should combine just fine, yes |
| 19:44 | justin_smith | underplank: not really, it gets created as part of making the other one |
| 19:44 | bordatoue | tolstoy: thanks |
| 19:44 | underplank | is there a way you can tell uberjar to just make the standalone one? |
| 19:44 | BengComps | I have been trying to run clojure on a macbook has anybody succedeed? |
| 19:44 | justin_smith | underplank: like I said, it uses it while making the other one, so the option isn't to make it or not, it is to delete it or not, but I don't think it has an option for deleting it |
| 19:45 | underplank | BengComps: doing it right now. what seems to be your problem? |
| 19:45 | underplank | justin_smith: fair enough |
| 19:45 | BengComps | any link should help, where to start underplank |
| 19:45 | underplank | I would start with Leiningen. |
| 19:45 | justin_smith | BengComps: I think more people use osx for clojure than use windows |
| 19:46 | underplank | http://leiningen.org/ |
| 19:46 | justin_smith | lein is just a single download |
| 19:46 | justin_smith | very easy to use |
| 19:47 | BengComps | Thanks a lot guys |
| 19:47 | BengComps | Right on it |
| 19:47 | BengComps | just for curious sake does any of you uses latex? and if so how do you run texmaker on yosemite |
| 19:48 | justin_smith | someone was talking the other day about using org mode with latex and clojure together, but now I forget who it was |
| 19:49 | justin_smith | to do inline calculations, then charts and such from those calculations |
| 19:50 | BengComps | I have been using sharelatex.com but the online compiler crashes as soon as my documents has a lot media content in it |
| 19:52 | tolstoy | Ugh. It's difficult when folks use that graphic for Clojars dependency declarations. Can't select it! |
| 19:53 | justin_smith | tolstoy: that graphic is svg, svg is selectable |
| 19:53 | justin_smith | unless your browser is bad |
| 19:53 | tolstoy | I mean select it to paste it into Emacs. Hm. |
| 19:53 | justin_smith | tolstoy: right, it has real text in it |
| 19:54 | justin_smith | which you can Control+F for, or drag and select etc. |
| 19:54 | tolstoy | Yeah, it doesn't select for me. Safari loses? |
| 19:54 | justin_smith | I guess |
| 19:54 | justin_smith | hmm, one moment |
| 19:54 | justin_smith | tolstoy: remind me again one of the libs that uses that? |
| 19:54 | tolstoy | https://github.com/xsc/pandect ;; works for you there? |
| 19:55 | justin_smith | tolstoy: oh man, they way they embed it, it doesn't actually work |
| 19:55 | underplank | oh.. yeah that does seem anoying… is it because it comes straight from clojars? |
| 19:55 | underplank | so they dont have to update it? |
| 19:55 | tolstoy | No idea. |
| 19:55 | justin_smith | so I guess I should have said, svg text *should be* selectable |
| 19:56 | andyf | BengComps: There is a program called MacPorts that lets you install many different open source software packages, including latex. |
| 19:56 | underplank | Yeah. Its way of getting the latest version from clojars in the README and not have it out of date. |
| 19:56 | underplank | the md is [](http://clojars.org/pandect) |
| 19:56 | justin_smith | underplank: I knew it - you have to open the svg itself |
| 19:56 | justin_smith | https://camo.githubusercontent.com/63193fb6d7a245fd83b93937eed69353e06121f9/687474703a2f2f636c6f6a6172732e6f72672f70616e646563742f6c61746573742d76657273696f6e2e737667 |
| 19:57 | justin_smith | on the svg itself (accessed via right-click / open image in new tab) you can select the text |
| 19:57 | underplank | annoying.. |
| 19:57 | justin_smith | the problem is the way it is embedded, or the css in the parent document |
| 19:57 | tolstoy | You might as well just click on it, _then_ select the text. |
| 19:58 | justin_smith | right, but theoretically, since it is an svg, the text should be selectable |
| 19:58 | underplank | I guess the problem is making it a cross realm query… svg/images makes that easy. I guess if you wanted it as text you would have to soemthing else? jsonp or something? |
| 19:58 | justin_smith | but it isn't when inside that page, for some weird reason |
| 19:59 | justin_smith | underplank: yeah, I think that's part of it |
| 20:00 | justin_smith | I've always thought svg was underutilized and poorly exploited. The fact that you can exactly control the display, and it has selectable / copyable text is awesome |
| 20:00 | justin_smith | and under utilized |
| 20:00 | Frozenlock | It is. |
| 20:01 | Frozenlock | A project in cljs almost entirely in SVG https://labs.hvac.io/editor |
| 20:01 | justin_smith | cool, is that yours? |
| 20:01 | Frozenlock | yes |
| 20:01 | justin_smith | very nice! |
| 20:02 | tolstoy | That's really great. |
| 20:03 | justin_smith | lol |
| 20:03 | Frozenlock | justin_smith: Make sure to save it :-p |
| 20:04 | justin_smith | nice! |
| 20:05 | justin_smith | Frozenlock: was this a student thing, or a learning exercise? |
| 20:06 | Frozenlock | No, it's a project in development for my own needs. |
| 20:07 | justin_smith | cool |
| 20:11 | Frozenlock | (thus the rough edges) |
| 21:16 | catern | I have two infinite sequences |
| 21:16 | catern | I'd like to merge them, sorted by some comparator |
| 21:16 | catern | idiom? |
| 21:17 | TEttinger | I'm not sure if there's a way to sort a literal infinite sequence |
| 21:17 | catern | ah... good point... |
| 21:17 | andyf | I think catern means to merge them, producing a new infinite sequence? |
| 21:17 | catern | both of them are increasing sequences, though |
| 21:18 | andyf | I assumed you meant that the two were already sorted |
| 21:18 | catern | andyf: yes |
| 21:18 | catern | (temporarily forgot that after TEttinger's remark) |
| 21:18 | andyf | Merging them is possible. Can hack up a few lines of code here in a minute or three. No 1-line version I?m aware of |
| 21:18 | justin_smith | I know I've written this function before |
| 21:18 | justin_smith | and yeah, it was like a three liner |
| 21:19 | catern | three lines?! so verbose ;_; |
| 21:19 | TEttinger | ha |
| 21:19 | ccallebs | Is there a specific clojure beginners channel? |
| 21:19 | catern | #clojure-beginners |
| 21:19 | ccallebs | Ah, nevermind. I see it in the MOTD :) |
| 21:19 | catern | :) |
| 21:19 | ccallebs | Sorry, using a new IRC client and it wasn't immediately obvious. |
| 21:20 | catern | so, what would be the general approach to this merging? |
| 21:20 | catern | i mean |
| 21:20 | catern | I guess I know how to actually manually do this |
| 21:25 | sobel | why might timbre prevent a (fairly simple) program from exiting after its -main returns? |
| 21:25 | sobel | i've narrowed it down to: if i call (info) to make a log entry it doesn't exit. |
| 21:25 | sobel | if i don't make any log calls, it exits normally |
| 21:29 | justin_smith | ,(defn merge-sort [& seqs] (let [[[s0 & ss :as s] & seqs] (sort-by first (remove empty? seqs))] (lazy-seq (when (seq s) (cons s0 (apply merge-sort (cons ss seqs))))))) |
| 21:29 | clojurebot | #'sandbox/merge-sort |
| 21:29 | freddd_ | hey, again. i'm mapping over a list of files and would like to ultimately have a hash-map output with keys of the file names and values of the file contents. the solution i'm using seems lacking, and i'm wondering if anyone can offer some advice. mock code here: https://www.refheap.com/99243 |
| 21:29 | justin_smith | ,(merge-sort [0 1 2 3] [2 3 4] [-1 8 9 10]) |
| 21:29 | clojurebot | (-1 0 1 2 2 ...) |
| 21:30 | justin_smith | that's a mouthful of a destructuring |
| 21:31 | justin_smith | freddd_: you can use a vector instead of a hash-map for each one |
| 21:31 | justin_smith | freddd_: also, unless the keywords ever get used as literals in your code, it's better to just keep the strings as keys |
| 21:32 | justin_smith | sobel: try shutdown-agents |
| 21:32 | justin_smith | sobel: timbre might be using futures in order to avoid slowing down your code, since the logging produces no result your code needs |
| 21:32 | sobel | likely |
| 21:33 | justin_smith | sobel: it uses a future to get your hostname? not seeing other usages of future (other than logged-future, but I don't see any calls to that within its codebase) |
| 21:33 | freddd_ | justin_smith: isn't it still wasteful to create some inner collection that just gets flattened anyway? |
| 21:33 | justin_smith | anyway, all it takes is one call to future, then you need shutdown-agents |
| 21:34 | justin_smith | freddd_: the vector does not get flattened, because maps are made of two element vectors |
| 21:34 | sobel | justin_smith: thanks! that was far from clear to me. |
| 21:34 | justin_smith | and it can reuse the underlying datastructure (and does) |
| 21:35 | sobel | (and fixed my not-exiting problem, which i diligently chased from the stupid start of my processing chain) |
| 21:35 | justin_smith | freddd_: eg. ##(seq {:a 0 :b 1 :c 2}) |
| 21:35 | lazybot | ⇒ ([:c 2] [:b 1] [:a 0]) |
| 21:35 | sobel | s/diligently/stupidly |
| 21:35 | justin_smith | freddd_: those are the vectors I am talking about ^ |
| 21:35 | justin_smith | catern: oh, I forgot to highlight you when I pasted my merge-sort above |
| 21:36 | justin_smith | anyway, my merge-sort, which I pasted above, should do what you want |
| 21:36 | catern | justin_smith: yeah, I got it. I actually need to do sort-by kind of thing, so I'm modifying it |
| 21:36 | andyf | catern: Not nearly as pretty, and only works for 2 input sequences, but https://gist.github.com/jafingerhut/41e9ceb258d1ddf05eff |
| 21:37 | sobel | one more i haven't solved: i got timbre logging to a file but can't disable console output |
| 21:37 | justin_smith | andyf: that one is likely faster for the two list case than mine though |
| 21:37 | justin_smith | sobel: did you remove the console logger from its config atom? |
| 21:38 | sobel | justin_smith: not finding the key easily |
| 21:39 | sobel | aha :standard-out |
| 21:39 | justin_smith | sobel: it will be one of the :appenders |
| 21:39 | justin_smith | yeah |
| 21:40 | justin_smith | you should be able to dissoc that from the atom |
| 21:40 | catern | hmmm |
| 21:41 | sobel | it has a convenienc fcn |
| 21:41 | sobel | ok, sweet. shoulda come here sooner. haha. and had more confidence in the WORKING part of my code. |
| 21:42 | justin_smith | heh |
| 21:45 | catern | thanks again justin_smith |
| 21:46 | catern | yours is cool too andyf :) |
| 21:47 | morgan_ | hi all, i have a question about clojure's data representation philosophy. when i have some data, do i annotate it with type information even when this is unnecessary for the logic of the program? |
| 21:48 | morgan_ | simple example. to represent a point in the plane, should i just use a pair [x y] of doubles |
| 21:48 | morgan_ | or should i define a record, like |
| 21:48 | morgan_ | (defrecord point [x y]) |
| 21:52 | justin_smith | morgan_: a record is an extended hash map, so it is easy to code with maps {:x 1 :y 3.3} and upgrade to a record later if you need it |
| 21:53 | justin_smith | also, records are idiomatically capitalized, because they generate a Class |
| 21:54 | morgan_ | justin_smith: thanks! so let me see if this correct: in idiomatic clojure, you wouldn't define a record for a type of data unless you needed the special capabilities of records (ability to implement protocols, etc., i assume). correct? |
| 21:55 | justin_smith | right |
| 21:55 | catern | zomg |
| 21:55 | morgan_ | ok, thank you! |
| 21:55 | catern | why is dissoc even in the language when disj is there |
| 21:56 | justin_smith | because sets are not associative? |
| 21:56 | justin_smith | ,(disj {:a 0} :a) |
| 21:56 | clojurebot | #error{:cause "clojure.lang.PersistentArrayMap cannot be cast to clojure.lang.IPersistentSet", :via [{:type java.lang.ClassCastException, :message "clojure.lang.PersistentArrayMap cannot be cast to clojure.lang.IPersistentSet", :at [clojure.core$disj invoke "core.clj" 1453]}], :trace [[clojure.core$disj invoke "core.clj" 1453] [sandbox$eval25 invoke "NO_SOURCE_FILE" -1] [clojure.lang.Compiler eval... |
| 21:57 | catern | aw :( |
| 21:57 | justin_smith | ,(dissoc #{:a} :a) |
| 21:57 | clojurebot | #error{:cause "clojure.lang.PersistentHashSet cannot be cast to clojure.lang.IPersistentMap", :via [{:type java.lang.ClassCastException, :message "clojure.lang.PersistentHashSet cannot be cast to clojure.lang.IPersistentMap", :at [clojure.lang.RT dissoc "RT.java" 806]}], :trace [[clojure.lang.RT dissoc "RT.java" 806] [clojure.core$dissoc invoke "core.clj" 1438] [sandbox$eval49 invoke "NO_SOURCE_FI... |
| 21:57 | catern | (i thought disj worked on maps) |
| 21:57 | justin_smith | catern: it would make sense to have one function that does both |
| 21:57 | catern | would it? |
| 21:57 | justin_smith | but probably too late for that design decision |
| 21:57 | justin_smith | catern: an inverse of conj, I guess |
| 21:58 | catern | yeah |
| 21:58 | catern | disj |
| 21:58 | justin_smith | catern: but conj works on maps |
| 21:58 | justin_smith | and lists |
| 21:58 | catern | yeah, disj should work on them |
| 21:58 | justin_smith | and vectors |
| 21:58 | TEttinger | and sets |
| 21:59 | justin_smith | TEttinger: well, luckily disj already works on one of these things |
| 21:59 | TEttinger | heh |
| 21:59 | TEttinger | ...futures? |
| 22:04 | justin_smith | ,(clojure.string/join "☃" "") |
| 22:04 | clojurebot | "☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☠|
| 22:04 | justin_smith | &(clojure.string/join "☃" "") |
| 22:04 | lazybot | ⇒ "☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☠|
| 22:04 | justin_smith | omghax |
| 22:05 | TEttinger | ,(count "") |
| 22:05 | clojurebot | 135 |
| 22:05 | TEttinger | I had to press left a lot to get through all them BOMs |
| 22:06 | TEttinger | ,(map int "") |
| 22:06 | clojurebot | (65279 65279 65279 65279 65279 ...) |
| 22:08 | TEttinger | ,(distinct "") |
| 22:08 | clojurebot | (\) |
| 22:11 | Seylerius | What was clojurebot's prefix for an inline eval? |
| 22:11 | TEttinger | lazybot uses ## |
| 22:11 | TEttinger | like this ##(+ 1 2 3) |
| 22:11 | lazybot | ⇒ 6 |
| 22:12 | justin_smith | Seylerius: but it needs some matching pair of brackets, otherwise talking about certain IRC channels would get annoying |
| 22:13 | catern | maybe it's ,,(inc 1) |
| 22:14 | sobel | how would i figure out, short of trying it, whether Robert.Hooke works in ClojureScript? |
| 22:16 | justin_smith | I forget whether clojurescript has vars (wiki still says no) but if not, then it won't work at runtime |
| 22:16 | justin_smith | afaik |
| 22:16 | justin_smith | https://github.com/clojure/clojurescript/wiki/Differences-from-Clojure#concurrent-programming |
| 22:17 | Seylerius | TEttinger, justin_smith: Thanks. I thought I saw something about clojurebot being able to do that too, but I can't find where someone mentioned it. |
| 22:39 | d9k_ | Hi! How to check in clojure module whether it's running from LightTable instarepl or from lein run? |
| 22:50 | justin_smith | d9k_: for one thin, lein run will call -main, the instarepl will not |
| 22:50 | justin_smith | unless you make it do that, of course |
| 22:52 | eraserhd | I'm having a huge brain fart.... how does one lexicographically compare vectors? |
| 22:52 | justin_smith | eraserhd: compare |
| 22:53 | justin_smith | ,(compare [:a :b :c] [:b :c :d]) |
| 22:53 | clojurebot | -1 |
| 22:54 | eraserhd | justin_smith: Thanks. |
| 23:00 | eraserhd | getting close to the next avi release: searching (`/`, `?`, `n`, `N`, `*`, and `#`). |
| 23:00 | eraserhd | I shall implement syntax highlighting as my next big project, I think. |
| 23:09 | d9k_ | How to check in clojure module whether it's running from LightTable instarepl or from lein run? [2] |
| 23:09 | justin_smith | d9k_: -main will get run from lein run, but probably not from an instarepl |
| 23:10 | justin_smith | d9k_: also, there is a :repl profile in lein, which you could combine with environ for some detection magic |
| 23:10 | justin_smith | d9k_: but your question makes me suspect that you have top level side effects, and you wouldn't need to worry about this if you moved them into function definitions |
| 23:16 | d9k_ | justin_smith, thanks, better than nothing |
| 23:17 | justin_smith | d9k_: unless you need to call -main from the instarepl, I would just set a var inside -main |
| 23:26 | tolstoy | How do you tell if a Datomic db has been initialized with the attributes you want? |
| 23:26 | tolstoy | If you transact all the attributes over again, you get a new "hash value" back when you (d/db conn). |
| 23:32 | tolstoy | I suppose I can use d/attribute to see if any of the attributes have been defined. |
](http://clojars.org/pandect/latest-version.svg)](http://clojars.org/pandect)){kind=link}