2011-05-04
| 00:03 | technomancy | amalloy: ah bugger; forgot it requires git master |
| 00:17 | technomancy | or not; it was a screw-up with the clucy lib needing ancient contrib |
| 00:17 | technomancy | anyway, should be ok now |
| 00:18 | brehaut | does anyone know whats involved with getting sqlite going with clojure? |
| 00:22 | dnolen | brehaut: not much, use the JDBC connector, setup your tables, querying w/ ClojureQL is fun. |
| 00:23 | dnolen | http://www.zentus.com/sqlitejdbc/ |
| 00:23 | brehaut | dnolen: thanks |
| 00:23 | brehaut | no maven thing for it? |
| 00:24 | dnolen | brehaut: supposedly it's here http://www.jarvana.com/jarvana/search?search_type=project&project=sqlite |
| 00:25 | brehaut | dnolen: thanks |
| 00:26 | dnolen | actually those are different according to this, http://www.xerial.org/trac/Xerial/wiki/SQLiteJDBC |
| 00:27 | hiredman | why not use derby or some other java embedded sql? |
| 00:27 | dnolen | brehaut: in my case when I was playing around, I actually used the first link and just dropped the .jar in lib. |
| 00:27 | technomancy | iiiiiinteresting: http://www.cloudbees.com/foss/ |
| 00:27 | hiredman | that you can get from maven |
| 00:27 | technomancy | sc |
| 00:27 | technomancy | cab |
| 00:27 | technomancy | geez emacs |
| 00:28 | brehaut | hiredman: ive never even heard of derby; i'll check it out |
| 00:28 | brehaut | dnolen: ok sure |
| 00:30 | mreynolds | brehaut: Also known by it's Brand "JavaDB" |
| 00:34 | semperos | if you know how to administer a db already, might as well use smth you know like sqlite |
| 00:35 | brehaut | i dislike administering all the databases i know ;) |
| 00:56 | semperos | $findfn [1 2 2 3 1 4 5 6 6] [1 2 3 4 5 6] |
| 00:56 | sexpbot | [clojure.core/distinct] |
| 01:02 | amalloy | &((comp seq sorted-set) [1 2 2 3 1 4 5 6 6]) |
| 01:02 | sexpbot | ⟹ ([1 2 2 3 1 4 5 6 6]) |
| 01:02 | amalloy | &((comp seq sorted-set) 1 2 2 3 1 4 5 6 6) |
| 01:02 | sexpbot | ⟹ (1 2 3 4 5 6) |
| 02:01 | matthias_ | hmm, why would lein repl open some random port? it used to always pick the same one |
| 02:07 | amalloy | if it did you could only have one lein repl open on your whole system at a time? |
| 02:54 | matthias_ | oh, i was using lein repl instead of lein swank. don't know why :| |
| 03:41 | fortxun | $seen rhickey |
| 03:41 | sexpbot | rhickey was last seen quitting 4 days and 12 hours ago. |
| 04:28 | Kototama | hello, does anyone now why analemma has been removed from clojars? |
| 06:17 | yayitswei | hello, I'm getting "#<CompilerException while parsing a block collection expected <block end>, but found ?" in my REPL |
| 06:17 | yayitswei | anyone know what that error means? |
| 06:17 | fliebel | yayitswei: That you have a syntax error. |
| 06:19 | yayitswei | thanks fliebel. just realized it was a syntax error in a yml file I'm parsing. how could I have debugged that? |
| 06:20 | fliebel | yayitswei: By printing the stacktrace and looking for files that belong to your project. |
| 06:20 | fliebel | It'll tell you line numbers. |
| 06:20 | yayitswei | ahh thanks. how do i print the stacktrace in the repl? |
| 06:20 | fliebel | (.printStackTrace *e) |
| 06:22 | fliebel | devn: ping |
| 06:24 | yayitswei | got it. thx |
| 08:35 | fortxun | $seen rhickey |
| 08:35 | sexpbot | rhickey was last seen quitting 4 days and 17 hours ago. |
| 08:47 | clgv | $seen fortxun |
| 08:47 | sexpbot | fortxun was last seen talking on #clojure 12 minutes and 19 seconds ago. |
| 08:58 | @chouser | fortxun: you could try the google group or his email address |
| 08:59 | @chouser | or I suppose knock on his door, if you know where that is. |
| 08:59 | solussd | Anybody here made it through the 'conjure' tutorial? I am stuck on 'scaffolding', getting: JdbcSQLException: Wrong user name or password |
| 08:59 | fortxun | chouser, what is his email address? |
| 09:01 | chouser | fortxun: I would think you could find it on here somewhere: http://groups.google.com/group/clojure-dev |
| 09:09 | rak85 | hi guys |
| 09:09 | rak85 | i have a list of vectors in the following format -> '([2 4] [4 343] [5 500]) |
| 09:10 | rak85 | and I want to compare the second index from each vector and return the maximum |
| 09:10 | rak85 | in this case, i would return 500 |
| 09:10 | rak85 | how would you do that? |
| 09:10 | rak85 | i'm using reduce... |
| 09:11 | rak85 | any other approach? |
| 09:11 | ejackson | ,(map (comp max second) [[2 4] [4 343] [5 500]]) |
| 09:11 | clojurebot | (4 343 500) |
| 09:11 | rak85 | hmmm... |
| 09:11 | ejackson | no quite.... |
| 09:11 | ejackson | ,(max (map second [[2 4] [4 343] [5 500]])) |
| 09:11 | clojurebot | (4 343 500) |
| 09:11 | ejackson | lol.... i should do this in my own repl.... |
| 09:11 | ejackson | ,(apply max (map second [[2 4] [4 343] [5 500]])) |
| 09:11 | clojurebot | 500 |
| 09:12 | rak85 | i forgot to tell you something... |
| 09:12 | ejackson | now you sound like my clients |
| 09:12 | rak85 | i may have nil values in the second index... |
| 09:12 | rak85 | hahaha |
| 09:12 | ejackson | throw a map identity in there |
| 09:13 | ejackson | filter identity, rather |
| 09:13 | rak85 | hmmm..thanks |
| 09:13 | ejackson | ,(apply max (filter identity (map second [[2 4] [4 _] [5 500]]))) |
| 09:13 | clojurebot | java.lang.Exception: Unable to resolve symbol: _ in this context |
| 09:14 | ejackson | ,(apply max (filter identity (map second [[2 4] [4 nil] [5 500]]))) |
| 09:14 | clojurebot | 500 |
| 09:14 | rak85 | ejackson: thanks! |
| 09:14 | ejackson | np |
| 09:17 | raek | ,(reduce max (for [[x y] [[2 4] [4 343] [5 500]] :when y] y)) ;; Another approach |
| 09:17 | clojurebot | 500 |
| 09:44 | clgv | &(->> [[2 4] [4 343] [5 500]] (map second) (reduce max)) |
| 09:44 | sexpbot | ⟹ 500 |
| 09:51 | warpy | can someone point me to clojure 1.2.1 changelog? |
| 09:51 | opqdonut_ | https://github.com/clojure/clojure/blob/1.2.x/changes.txt |
| 09:52 | warpy | thanks opqdonut_ |
| 09:52 | opqdonut_ | np |
| 10:06 | __name__ | &(->> [[2 4] [4 343] [5 500]] (map second) ma) |
| 10:06 | sexpbot | java.lang.Exception: Unable to resolve symbol: ma in this context |
| 10:06 | __name__ | &(->> [[2 4] [4 343] [5 500]] (map second) max) |
| 10:06 | sexpbot | ⟹ (4 343 500) |
| 10:06 | __name__ | &(->> [[2 4] [4 343] [5 500]] (map second) (apply max)) |
| 10:06 | sexpbot | ⟹ 500 |
| 10:06 | __name__ | clgv: Is that not nicer? |
| 10:08 | clgv | __name__: why should it be? the question is: nicer to which "measure"? ;) |
| 10:08 | __name__ | clgv: reduce seems a bit redundant because it is more function calls. |
| 10:08 | clgv | on a pure text comparison it's just recue exchange with apply ;) |
| 10:09 | raek | the vararg version of max is implemented with reduce... |
| 10:09 | clgv | __name__: max with more the two arguments is also more than one function call ;) |
| 11:59 | gfrlog | so I'm using compojure and ring... |
| 11:59 | gfrlog | and I'm testing the server by sending it requests manually, since the server is a function... |
| 11:59 | gfrlog | but I'm having trouble interacting with it the same way jetty does |
| 12:00 | manutter | what kind of trouble? |
| 12:00 | gfrlog | specifically wrt form and route params |
| 12:00 | KirinDave_ | gfrlog: Well yeah, jetty is a ton of environment and side effects underneath your code. |
| 12:00 | KirinDave_ | gfrlog: The way I did it was I put a capture handler that dumped out ALL the data structures for a request to file. |
| 12:00 | KirinDave_ | gfrlog: I'd reconstitute those for my tests |
| 12:01 | gfrlog | hmm |
| 12:01 | KirinDave_ | gfrlog: The only trick there is you need to manually handle the body of the post, which is an input stream. |
| 12:01 | gfrlog | what I was doing was just passing in the form-params as a map |
| 12:01 | mattmitchell | is the a nice way to create a hash-map from a 2 vectors, where one vector contains the keys and the other the values? |
| 12:02 | KirinDave_ | mattmitchell: zipmap |
| 12:02 | mattmitchell | KirinDave_: excellent thanks! |
| 12:02 | KirinDave_ | mattmitchell: Is there some sort of challenge or homework or something? You're probably the 5th person to ask that in 24 hours |
| 12:02 | gfrlog | KirinDave_: so I was asking because I was hoping there was a not-messy way to figure this out. It sounds like you're saying it's just gonna be messy? |
| 12:02 | KirinDave_ | gfrlog: It shouldn't take too long. |
| 12:03 | mattmitchell | KirinDave_: no :) guess it's just a common problem to solve |
| 12:03 | KirinDave_ | mattmitchell: Well, zipmap is the answer :) |
| 12:03 | gfrlog | KirinDave_: I wouldn't expect so, it just makes me feel gross and sticky :) |
| 12:04 | manutter | gfrlog: think of it as building a mock-jetty for testing purposes :) |
| 12:04 | KirinDave_ | It's weird to me that there is zipmap but no zipwith. |
| 12:04 | gfrlog | ,(zipmap (iterate inc 0) (iterate inc 1)) |
| 12:04 | clojurebot | Execution Timed Out |
| 12:04 | gfrlog | manutter: I think I was hoping that the code I was missing was in compojure somewhere rather than in jetty |
| 12:04 | KirinDave_ | Ahh yes |
| 12:04 | gfrlog | or ring |
| 12:04 | KirinDave_ | gfrlog: The answer is no. |
| 12:04 | KirinDave_ | There is no default request content. |
| 12:05 | KirinDave_ | But you'll see |
| 12:05 | KirinDave_ | Once you have a few for inspection, you'll find modifying them is relatively easy. |
| 12:05 | gfrlog | KirinDave_: it's not all the boilerplate that I'm missing, it's the method of assembling the input |
| 12:05 | gfrlog | e.g., the form parameters are duplicated in the :form-params key and the :params key |
| 12:05 | KirinDave_ | "assembling the input"? I don't understand |
| 12:05 | gfrlog | where :params also has the route parameters |
| 12:05 | KirinDave_ | Well that's because you put a handler there I assume. |
| 12:05 | gfrlog | and maybe the query parameters as well |
| 12:05 | manutter | you mean to turn a map into an encoded string, ya? |
| 12:06 | KirinDave_ | There is like a unified params handler. |
| 12:06 | gfrlog | KirinDave_: yes, the unified one works with browser/jetty |
| 12:06 | KirinDave_ | If you're using that, params is assembled. |
| 12:06 | gfrlog | but when I call the app directly, that doesn't get done |
| 12:06 | KirinDave_ | I'm of the opinion params is dangerous, but also devilishly convenient. :) |
| 12:06 | KirinDave_ | Indeed. |
| 12:06 | KirinDave_ | Sorry, my greyhound needs walking. |
| 12:06 | raek | gfrlog: those are usually parsed from the :url or :body by some ring middleware, but maybe compojure handles that automatically |
| 12:06 | gfrlog | you can't destructure as easily with form-params though, because it has string keys |
| 12:06 | KirinDave_ | He states this by draping his body across my hands. |
| 12:07 | manutter | lol |
| 12:07 | KirinDave_ | And I have to support his head with my wrists. Progressively increasingly difficult to type. |
| 12:07 | gfrlog | maybe I will try encoding my form params into the body and see if it gets handled that way |
| 12:07 | gfrlog | KirinDave_: okay have fun :) |
| 12:07 | gfrlog | thanks for the help both of you. lunch calls. |
| 12:07 | KirinDave_ | gfrlog: Why not just capture some request structures? Seriously. It takes all of 5 minutes. Use curl to hit a random endpoint with the kind of requests you want. |
| 12:08 | KirinDave_ | Then you've got lightning in the bottle. |
| 12:08 | gfrlog | KirinDave_: I may not have explained myself well enough, because I don't think that's really the issue. But I think I understand things better now, so it'll be okay. |
| 12:09 | dnolen | raek: I fixed those issues from yesterday (I hope) |
| 12:10 | devn | fliebel: im here |
| 12:11 | devn | fliebel: sorry, would have gotten back to you yesterday but was too busy flopping around on the ground (seizure) |
| 12:14 | Ramblurr | in enlive, what is the proper way to do a conditional transform? e.g., [:selector] (if condition (content "success!") (some_other_transform)) |
| 12:19 | raek | Ramblurr: the form to the right should be a function that takes a node and returns a node, so you can do it like this (fn [match] (if (pred? match) ((content "success!") match) ((some-other-transform) match)), I believe |
| 12:19 | raek | oh, the enlive readme explains it even better: "Transformations (the right-hand parts of rules) are now plain old closures. |
| 12:20 | raek | These functions take one arg (the selected node) and return nil, another node |
| 12:20 | raek | or a collection of nodes." |
| 12:20 | Ramblurr | ah i saw that, it didn't click. so using a closure should work |
| 12:20 | raek | so 'content' is a function that returns a transform |
| 12:22 | raek | dnolen: ok, nice. then I don't have to move the issue to the correct location... :) |
| 12:23 | Ramblurr | raek: you mean 'content' is a function that returns a transform function |
| 12:23 | Ramblurr | ..right? or a col of nodes? |
| 12:23 | Ramblurr | gonna try it out.. |
| 12:23 | raek | yes, a transform function.. |
| 12:24 | raek | sorry |
| 12:27 | Ramblurr | raek: ah ive got it working |
| 12:27 | Ramblurr | i was trying to pass match to 'content' itself |
| 12:27 | Ramblurr | cause i thought it was content that returned the col of nodes |
| 12:51 | noidi_ | Why doesn't the following form work in the ns macro: (:refer-clojure :exclude [*]) ? |
| 12:51 | noidi_ | Unable to resolve symbol: * in this context |
| 12:52 | noidi_ | excluding + doesn't work either |
| 12:52 | noidi_ | but e.g. excluding assoc works fine |
| 12:52 | noidi_ | is there something special about the math operators? |
| 12:52 | technomancy | noidi_: maybe because + and * are defined using definine? |
| 12:54 | technomancy | doesn't really make sense, but it's the only thing I can think of |
| 12:54 | noidi_ | ok, thanks |
| 12:56 | Raynes | noidi_: He meant definline |
| 12:57 | noidi_ | they don't seem to work as method names in a protocol either :P |
| 12:58 | noidi_ | so there's something special about them |
| 13:03 | chouser | (ns foo (:refer-clojure :exclude [*])) works for me |
| 13:04 | chouser | of course any attempt to use * after that results in Unable to resolve symbol: * in this context |
| 13:04 | noidi_ | I'm using 1.2.0 |
| 13:05 | chouser | hm, that was with 1.3-something. I'll try 1.2 |
| 13:05 | chouser | same result on 1.2 |
| 13:07 | noidi_ | oddly enough the error disappears if I remove the (:import) form from the ns call |
| 13:08 | noidi_ | so I have a syntax error that gets the macro confused |
| 13:08 | noidi_ | *maybe* |
| 13:10 | noidi_ | never mind that, the error was really caused by me calling * after excluding it |
| 13:10 | noidi_ | :P |
| 13:11 | noidi_ | I changed the protocol method's name from * since it didn't seem to work, but forgot to change the users back |
| 13:15 | noidi_ | yay, everything works now |
| 13:16 | noidi_ | I just had my REPL in a weird state after all the experimentation with the names |
| 13:16 | noidi_ | sorry about that :) |
| 13:22 | chouser | noidi_: not a problem. glad you figured it out |
| 13:28 | lawfulfalafel | okay I am trying to do what it says on this page: http://dev.clojure.org/display/doc/Getting+Started+with+Emacs and I can't get the M-x package-list-packages command to work |
| 13:28 | lawfulfalafel | I think the icicles mode is overriding it |
| 13:28 | dpritchett | lawfulfalafel, are you sure you have elpa installed? |
| 13:29 | lawfulfalafel | elpa? |
| 13:29 | clojurebot | elpa is a package manager for Emacs: http://tromey.com/elpa |
| 13:30 | gfrlog | any quick/easy way to encode a map as POST params? |
| 13:30 | lawfulfalafel | okay I ran that install script |
| 13:31 | amalloy | KirinDave_: isn't zipwith just a special case of map? ##(map + (range) (range 1 10)) |
| 13:31 | sexpbot | ⟹ (1 3 5 7 9 11 13 15 17) |
| 13:31 | gfrlog | ,(doc zipwith) |
| 13:31 | clojurebot | No entiendo |
| 13:32 | amalloy | gfrlog: he means the haskell/scala/whatever zipwith |
| 13:32 | gfrlog | dangit I don't know those languages |
| 13:32 | gfrlog | is it my map-from-fn idea? |
| 13:32 | KirinDave_ | amalloy: It's just weird to have zipmap but not zipwith. :) |
| 13:32 | lawfulfalafel | dpritchett: thanks |
| 13:32 | amalloy | gfrlog: i don't know what idea you're talking about, but no |
| 13:33 | manutter | gfrlog: I wonder if clj-http has any form-encoding functions |
| 13:33 | manutter | (assuming I have the name right) |
| 13:33 | dpritchett | np lawfulfalafel |
| 13:33 | KirinDave_ | amalloy: I'm sure zipmap is just an implementation optimization for a common case. |
| 13:34 | amalloy | KirinDave_: i doubt it's an implementation optimization |
| 13:34 | manutter | gfrlog: https://github.com/mmcgrana/clj-http |
| 13:34 | amalloy | &(into {} (map vector (range) (range 1 10))) |
| 13:34 | sexpbot | ⟹ {0 1, 1 2, 2 3, 3 4, 4 5, 5 6, 6 7, 7 8, 8 9} |
| 13:34 | gfrlog | KirinDave_: amalloy: it's not |
| 13:34 | KirinDave_ | Weird. |
| 13:35 | gfrlog | the source just uses (loop) and goes one by one |
| 13:35 | amalloy | KirinDave_: zipmap is doing a different thing than zipwith would |
| 13:35 | KirinDave_ | amalloy: Because it's a reduce. |
| 13:35 | amalloy | right |
| 13:35 | KirinDave_ | Still you could do zipmap in terms of zipwith. |
| 13:35 | amalloy | obviously. i just did |
| 13:36 | KirinDave_ | I guess zipmap is just the belle of the zip family. :) |
| 13:36 | KirinDave_ | So in demand, so famous, so glamorous. |
| 13:36 | dakrone | ,(apply str (interpose "&" (map (comp (partial apply str) (partial interpose "=")) {"a" "foo" "b" "bar"}))) |
| 13:36 | clojurebot | "a=foo&b=bar" |
| 13:36 | manutter | heh |
| 13:37 | manutter | I think you also need to map <space> to + and special chars to, what is it %xx; format? |
| 13:37 | gfrlog | dakrone: the escaping is the hard part |
| 13:38 | dakrone | that's true |
| 13:38 | amalloy | dakrone: i like point-free as much as the next guy, but that partial stuff is kinda overdoing it ##(clojure.string/join "&" (map (fn [[k v]] (str k "=" v)) {"a" "foo" "b" "bar"})) |
| 13:38 | sexpbot | ⟹ "a=foo&b=bar" |
| 13:39 | amalloy | anyway this is a wheel that has been invented a lot of times. that said, i was looking for it the other day and didn't find any convenient clojure wrappers for it |
| 13:39 | dakrone | :) |
| 13:39 | gfrlog | amalloy: what is point-free? |
| 13:40 | amalloy | $google haskell point free |
| 13:40 | sexpbot | First out of 93500 results is: Pointfree - HaskellWiki |
| 13:40 | sexpbot | http://www.haskell.org/haskellwiki/Pointfree |
| 13:40 | KirinDave_ | Point free is kinda pointless in lispy languages. |
| 13:40 | amalloy | gfrlog: tl;dr: building functions by gluing together functions without mentioning their arguments explicitly |
| 13:41 | amalloy | KirinDave_: i see no evidence of that |
| 13:41 | gfrlog | that is a good formalization of a concept that's been floating around my head for a while now |
| 13:41 | KirinDave_ | amalloy: It's more trouble than its worth, in my opinion. Lisp's rules for application don't facilitate it. |
| 13:41 | gfrlog | when I create point-free expressions, it gives me a bigger programming high and makes me not want to be around when a clojure-n00b has to support the code |
| 13:41 | gfrlog | so I do it all the time |
| 13:42 | KirinDave_ | amalloy: And also they tend to be longer in clojure because you gotta use (partial ...) :) |
| 13:42 | gfrlog | yeah partial is way too long |
| 13:42 | S11001001 | gfrlog: well obviously you wouldn't have made any mistakes there, so said n00b knows to look elsewhere for problems |
| 13:42 | amalloy | KirinDave_: (defalias $ partial) :P |
| 13:42 | gfrlog | S11001001: yes, that is obvious |
| 13:42 | KirinDave_ | amalloy: Ha. |
| 13:42 | amalloy | and then add some greek letters |
| 13:42 | KirinDave_ | What could go wrong. |
| 13:43 | gfrlog | don't tell jquery |
| 13:43 | amalloy | KirinDave_: what could go wrong is you could end up with hiredman's code-rewriting macro |
| 13:43 | KirinDave_ | The upshot of a () wrapped partial call is that the binding order is way clearer |
| 13:43 | KirinDave_ | amalloy: What does that look like? |
| 13:43 | amalloy | it looks pretty cool tbh, but not a lot like lisp |
| 13:43 | KirinDave_ | I thought he super-hated macros. |
| 13:43 | amalloy | i can never find the dang thing |
| 13:43 | KirinDave_ | I mean, it takes a few doubletakes to resolve what x = (f .) . g |
| 13:44 | KirinDave_ | really does. |
| 13:44 | amalloy | KirinDave_: it takes me a few double takes to come up with (f .) . g when what it seems like i want is f . g |
| 13:44 | hiredman | https://github.com/hiredman/odds-and-ends/blob/master/functional.clj#L126 |
| 13:44 | amalloy | hooray, thanks hiredman |
| 13:45 | KirinDave_ | hiredman: Whoah |
| 13:45 | KirinDave_ | hiredman: Ha. At first i was like, "Clj doesn't have flip in the stdlib... ohhhh wait line 55) |
| 13:45 | hiredman | uncurry is still completely the opposite of the name |
| 13:46 | KirinDave_ | :) |
| 13:46 | gfrlog | add an ignored argument? |
| 13:51 | mattmitchell | i know there is a "some" which will return the non-false return value from the predict. Is there anything that will return the item in the seq when the predicate returns true? |
| 13:52 | amalloy | (comp first filter) |
| 13:53 | amalloy | or, write your predicate differently |
| 13:53 | mattmitchell | amalloy: aye that's pretty simple. thanks! |
| 13:53 | mattmitchell | yeah i came up with this: (defn detect [f col] (some #(if (f %1) %1) col)) |
| 13:53 | mreynolds | I've been tending towards using (partial) instead of constructing functions on-the-fly #(), are there any caveats or idiomatic reasons not to do this? |
| 13:54 | mattmitchell | then i can do (detect pred col) |
| 13:57 | amalloy | mattmitchell: fwiw, i prefer (defn detect [pred] (fn [x] (when (pred x) x))) (some (detect even?) coll) |
| 13:58 | amalloy | then you're not tied to "some" and can reuse that adjusted predicate anywhere (and believe me, it's useful in a lot of places) |
| 13:59 | amalloy | mreynolds: they're both fine. i like partial. but if you end up with more than a couple partials all in one place it tends to be hard to read, so split it up |
| 13:59 | mattmitchell | amalloy: ahh cool |
| 13:59 | mreynolds | amalloy: Gotcha, yeah, I've been using that same pattern for readability. Thanks. |
| 14:00 | chouser | mreynolds: fwiw, each literal #() in your code at compile time produces a new class, while (partial) does not. |
| 14:01 | chouser | on the other hand, a calling a fn created with partial may require an extra method call or two at runtime, but then again those are probably inlined away in any loop tight enough to matter. |
| 14:01 | mreynolds | chouser: Ahhh, nice to know. Have you heard of people running into problems with too many classes? I guess I could see the JVM choking on having to manage a few hundred thousand |
| 14:02 | chouser | I think the biggest real problem with lots of classes is the startup overhead for the JVM to validate all of them |
| 14:03 | chouser | but again, that's probably something to be improved in a more systemic way, rather than replacing each of your uses of #() with partial. :-P |
| 14:04 | mreynolds | hehe, yeah. So far, I haven't seen it as a problem anyway. I'm using Clojure for services and server-side stuff where startup time hasn't been an issue, neither class GC/management. |
| 14:06 | amalloy | chouser: surprisingly, i never realized partial doesn't need a new class for each use. it's clear if you think about it, but i never did |
| 14:07 | amalloy | i just saw core$partial$fn__3682 and was like, yep, there goes another anonymous class |
| 14:07 | chouser | amalloy: I mention it alot because it surprised me too when I first realized that difference |
| 14:07 | chouser | alot |
| 14:07 | amalloy | haha |
| 14:08 | chouser | alots are always surprising me as they lumber on by |
| 14:09 | gfrlog | man I'm having this horrible time trying to test my compojure app without making my app code really clunky |
| 14:09 | amalloy | $dict alot |
| 14:09 | sexpbot | amalloy: noun: common misspelling of a lot. |
| 14:09 | amalloy | nice |
| 14:09 | gfrlog | $dict cleave |
| 14:09 | sexpbot | gfrlog: verb-transitive: To split with or as if with a sharp instrument. See Synonyms at tear1. |
| 14:09 | amalloy | i was hoping it was some real thing, when you said they lumber on by |
| 14:10 | gfrlog | oh is somebody referencing hyperbole and a half? |
| 14:10 | gfrlog | $google hyperbole half alot |
| 14:10 | sexpbot | First out of 74500 results is: Hyperbole and a Half: The Alot is Better Than You at Everything |
| 14:10 | sexpbot | http://hyperboleandahalf.blogspot.com/2010/04/alot-is-better-than-you-at-everything.html |
| 14:10 | mreynolds | I should probably look at the return codes from functions more often to understand the impl better |
| 14:10 | manutter | sexpbot scares me on a daily basis. |
| 14:10 | semperos | lol |
| 14:11 | amalloy | mreynolds: the return codes? |
| 14:12 | gfrlog | so compojure passes in four keys -- query-params, form-params, route-params, and params |
| 14:12 | gfrlog | the easiest way to do bindings in compojure is straight out of params |
| 14:12 | gfrlog | but if I call the app directly in my test, I can't set params up the same way that jetty? does it |
| 14:12 | gfrlog | because I don't have the route-params at that point |
| 14:12 | gfrlog | only the form params |
| 14:13 | gfrlog | binding to the form-params specifically in the app code is tough because it uses strings for keys |
| 14:13 | gfrlog | when they are merged into the main params map, they're converted to keywords |
| 14:14 | amalloy | oh my god i am in love with that article. thanks, chouser and gfrlog, for bringing it to my attention |
| 14:14 | manutter | gfrlog: by "jetty" do you mean the ring module that wraps the calls to the Jetty servlet? |
| 14:14 | gfrlog | manutter: I could mean anything |
| 14:14 | gfrlog | thus the question mark |
| 14:15 | gfrlog | I don't understand at what point the things happen |
| 14:15 | gfrlog | but apperently the app I get from compojure that I call from my tests does not include the functionality |
| 14:15 | amalloy | gfrlog: isn't it ring.middleware.params/wrap-params that turns the query string into a keyword map? |
| 14:15 | gfrlog | amalloy: you're welcome. I'm honored to have contributed |
| 14:15 | amalloy | or a string map |
| 14:16 | gfrlog | amalloy: maybe? would you expect that to be included in the app then? |
| 14:16 | amalloy | you can wrap your handler with that manually when you're testing |
| 14:16 | amalloy | gfrlog: by app, you mean...? |
| 14:16 | gfrlog | I mean the function returned from...hmm |
| 14:16 | gfrlog | compojure.handler/api |
| 14:17 | gfrlog | I think |
| 14:17 | gfrlog | is what I'm passing the routes to |
| 14:17 | gfrlog | so my app-function I am testing by calling it as such: |
| 14:17 | mreynolds | amalloy: Er, sorry, the values as printed by pprint :) AKA function and object ids |
| 14:18 | gfrlog | (app {:request-method ..., :uri ..., :body "foo=bar"}) |
| 14:18 | amalloy | gfrlog: i think so, but you can look at the compojure source and see |
| 14:19 | amalloy | https://github.com/weavejester/compojure/blob/master/src/compojure/handler.clj#L15 |
| 14:23 | gfrlog | amalloy: my app is seeing the :body attr exactly as I passed it in, and the :form-params attribute an empty map |
| 14:24 | gfrlog | amalloy: I guess this is enough to dig into it on my own, thanks |
| 14:24 | amalloy | gfrlog: earlier you said you only had the form params, not the route params. which is it you have and which do you wish you had? |
| 14:24 | gfrlog | I as the tester only have the form params. so I want to pass them in either as a :body or as :form-params, and I want them to be merged into :params along with the :route-params |
| 14:25 | gfrlog | when I try this, my app only sees the :route-params included in the :params map |
| 14:26 | gfrlog | (on a side note, the alot article is apparently the top google hit for "alot") |
| 14:27 | gfrlog | looking at the wrap-params method in the ring middleware... |
| 14:32 | amalloy | gfrlog: try just having your app output the request map it receives as the response map. then you can see exactly what it's getting and try fixing that |
| 14:32 | gfrlog | I can see exactly what it's getting |
| 14:32 | gfrlog | I am logging the request map |
| 14:33 | gfrlog | the naive "fix" would be to just set "params" myself |
| 14:33 | gfrlog | but then it would not include the route-params |
| 14:35 | gfrlog | maybe I need to set the content type |
| 14:35 | gfrlog | and then it will decode the body for me |
| 14:38 | gfrlog | it will decode it for me if I make it slurpable |
| 14:38 | gfrlog | how can I use $findfn to figure out how to make (slurp (f "foo")) return "foo"? |
| 14:38 | amalloy | $findarg slurp (% "foo") "foo" |
| 14:38 | sexpbot | [] |
| 14:38 | amalloy | no, i guess not |
| 14:39 | amalloy | still, you probably want ##(slurp (StringReader. "foo")) |
| 14:39 | sexpbot | java.lang.IllegalArgumentException: Unable to resolve classname: StringReader |
| 14:39 | amalloy | still, you probably want ##(slurp (java.io.StringReader. "foo")) |
| 14:39 | sexpbot | ⟹ "foo" |
| 14:39 | gfrlog | very good :) |
| 14:39 | gfrlog | thx |
| 14:41 | gfrlog | amalloy: I think all my problems just went away |
| 14:41 | amalloy | you're just not looking closely enough. additional problems abound |
| 14:41 | gfrlog | if it's small, does it really exist? |
| 14:42 | gfrlog | man that's full of interpretations? |
| 14:42 | gfrlog | s/?// |
| 14:48 | fliebel | devn: Still around? I was away all day. To bad about the seizure. |
| 15:00 | devn | fliebel: im here yes |
| 15:00 | devn | fliebel: (sort of) working remotely on some stuff in another terminal |
| 15:02 | fliebel | devn: Okay, just that your website seems to be down, and that I am interested in the status of your baker. And while I'm asking questions anyway, could you shed some light on the defn/devn confusion? |
| 15:12 | devn | fliebel: ive moved to devn permanently |
| 15:12 | devn | fliebel: defn, although i used it before doing clojure, started to look a little fanboyish |
| 15:12 | devn | fliebel: and also, my twitter handle was devn, so there is some unity now in my handles |
| 15:13 | devn | fliebel: the status of my baker is...slow... im working with luke vanderhart on a templating library that i want to use, so what i have done so far is not very exciting |
| 15:15 | fliebel | devn: What kind of templating? PHP style, or something exotic? |
| 15:19 | carllerche | What's the backing data structure for clojure's sorted set? |
| 15:20 | fliebel | carllerche: Some sort of tree for sure… I think the sorted ones use something binary. |
| 15:22 | devn | fliebel: google closure templates |
| 15:22 | devn | fliebel: https://github.com/levand/scribe |
| 15:24 | S11001001 | what should I call this? (fn [x & _] x) |
| 15:24 | amalloy | call it (comp first list) |
| 15:25 | S11001001 | ick |
| 15:25 | amalloy | carllerche: red-black tree |
| 15:25 | amalloy | i think |
| 15:25 | fliebel | amalloy: The sorted ones? |
| 15:25 | amalloy | fliebel: yeah |
| 15:26 | amalloy | or at least, that's what java's is. i suppose there is probably some changes necessary to handle persistence |
| 15:26 | fliebel | oh, nice. But I was right about the binary tree:) |
| 15:26 | amalloy | fliebel: definitely |
| 15:27 | gfrlog | (defn (comp first list) [x & _] x) |
| 15:28 | amalloy | yeah, it's still red/black |
| 15:28 | S11001001 | gfrlog: surely that only works with two-element lists whose firsts are 'setf |
| 15:34 | gfrlog | ,(doc setf) |
| 15:34 | clojurebot | Cool story bro. |
| 15:43 | fliebel | devn: Oh, good luck with that. Not my taste of templating. I am of the opinion that if you want to invent mini languages (for a dynamic repl environment), you might as well write it in the language itself. Therefore I prefer systems like Genshi, Enlive or even whatever became of compojure.html. |
| 15:45 | fliebel | Ah, Hiccup. I know it is terrible in a lot of ways, but I still love it. |
| 15:50 | thorwil | what are those terrible ways? |
| 15:51 | brehaut | thorwil: it doesnt escape text by default |
| 15:51 | fliebel | thorwil: It becomes very unmanageable for larger projects I've been told. |
| 15:52 | fliebel | And while every list can be expressed as an xml structure, not every xml structure can be expressed beautifully as a list. |
| 15:53 | thorwil | ah. so far i only stepped of enlive for a small snippet, where i simply used text to create tags |
| 15:54 | fliebel | thorwil: enlive != hiccup |
| 15:54 | thorwil | fliebel: yes, that one case just made me wonder if i should look into hiccup in addition |
| 15:55 | fliebel | thorwil: hiccup is for sure a lot less mind-bending than enlive for small snippets. |
| 15:56 | thorwil | my mind will already never be the same again ^^ |
| 15:57 | thorwil | though i still havn't found out how to substitute the tags around desired content with enlive |
| 16:07 | semperos | thorwil: what do you mean by "around"? code example? |
| 16:09 | thorwil | semperos: say i have <span>Previous</span> in my html template, and want to turn that into <a href="/foo">Previous</a> within an enlive html-snippet (depending on some condition) |
| 16:10 | semperos | understood |
| 16:15 | thorwil | gotta run, good night! |
| 16:18 | mattmitchell | someone wanna give me a hand with this? https://gist.github.com/955947 |
| 16:19 | semperos | if thorwil had waited two more seconds... |
| 16:19 | semperos | net.cgrand.enlive-html/substitute |
| 16:19 | mattmitchell | i'm close to whacking my head and going home if i can't figure this out! :( |
| 16:22 | Touqen | I think I'd just convert items into an actual tree datastructure, then figure out what I want to keep, and then re-flatten it |
| 16:22 | semperos | mattmitchell: could you explain the relationship between [1 2] and the other collection in your own words? |
| 16:23 | semperos | I see the result, but not sure I understand the exact relationship |
| 16:23 | amalloy | semperos: it's magic, duh |
| 16:23 | mattmitchell | semperos: sure. the [1 2] are id values |
| 16:23 | duncanm | is there a way to tell lein uberjar to *not* clear out the lib directory? |
| 16:23 | mattmitchell | amalloy: and yes it's magic! :) |
| 16:23 | duncanm | i have a jar that i want to use, and i really don't want to go about putting it in my maven repo |
| 16:23 | semperos | amalloy: more in the line of my stupid questions :) |
| 16:23 | amalloy | $google lein disable implicit clean |
| 16:23 | sexpbot | First out of 413 results is: how to use a jar not in a repository or project.clj? lein jar ... |
| 16:23 | mattmitchell | semperos: in otherwords... the [1 2] match the one of the :id values |
| 16:23 | sexpbot | http://groups.google.com/group/leiningen/browse_thread/thread/6cee060588793e97 |
| 16:24 | semperos | mattmitchell: ok, they correspond to the :id values |
| 16:24 | amalloy | srsly mattmitchell you're conflating too many concepts. separate out the "collapse parents and children" step from the "filter out IDs" step, i think |
| 16:24 | mattmitchell | semperos: yes exactly |
| 16:24 | semperos | duncanm: if you have the source, you can put it in the "checkouts" folder of your project layout, and leiningen will treat that like a jar |
| 16:25 | mattmitchell | amalloy: i'm not sure what you mean (sorry) |
| 16:25 | duncanm | it's a java project from netbeans, with its own dependencies |
| 16:25 | duncanm | semperos: so it's not that simple |
| 16:25 | semperos | it's not up in any repo? |
| 16:25 | manutter | mattmitchell: I just tried (into {} (map (juxt :id identity) items)) and got some results that might interest you |
| 16:25 | duncanm | sigh |
| 16:25 | amalloy | mattmitchell: you want (magic-fn) to not only combine multiple entities into one, but also filter out entities that don't meet some criteria |
| 16:26 | semperos | duncanm: sorry if these are repeat questions |
| 16:26 | amalloy | if you separate this enormous task into smaller tasks it will be easier to work with |
| 16:26 | duncanm | semperos: nah, i have it locally |
| 16:26 | mattmitchell | amalloy: i see what you mean, i will try to think of it that way |
| 16:26 | mattmitchell | manutter: awesome i'll see what that does! |
| 16:26 | amalloy | manutter: isn't that clojure.set/index? |
| 16:26 | amalloy | &(doc clojure.set/index) |
| 16:26 | sexpbot | ⟹ "([xrel ks]); Returns a map of the distinct values of ks in the xrel mapped to a set of the maps in xrel with the corresponding values of ks." |
| 16:27 | amalloy | not that i can ever read the #$% docs for clojure.set |
| 16:27 | manutter | amalloy: Is it? I'm such a noob! |
| 16:27 | semperos | duncanm: worst case scenario, you can always copy and paste whatever jars you need into your project's lib folder :) |
| 16:27 | duncanm | semperos: that's what i'm doing, and if i run uberjar, it first cleans up the lib folder |
| 16:27 | semperos | yep |
| 16:27 | semperos | it sure does |
| 16:27 | duncanm | and so it doesn't work ;-( |
| 16:27 | duncanm | i was hoping there's a no cleanup flag |
| 16:28 | manutter | duncanm: wonder if you could make a separate lein project consisting only of the dependencies, "lein jar" that, and then use the result in your real project? |
| 16:28 | manutter | or lein install it |
| 16:28 | mattmitchell | amalloy: aye, could you give me an example of what you mean by separating things a bit? |
| 16:28 | duncanm | manutter: still have to put it somewhere |
| 16:28 | duncanm | oh, maybe i can just lein install my java jar |
| 16:28 | duncanm | hmm |
| 16:28 | duncanm | sigh |
| 16:29 | mattmitchell | amalloy: you mean, for example a fn for only applying the parent code? and then another for attaching the parent items? |
| 16:29 | duncanm | i'm actually porting this clojure code from NetBeans/Ant to lein |
| 16:29 | duncanm | maybe i should just go back to ant, sucky as it may be |
| 16:30 | raek | duncanm: you can put :disable-deps-clean false in your project.clj |
| 16:30 | amalloy | (defn find-parent-of [all-data the-element] (first (filter (comp #{(:parent_id the-element)} :id) all-data)))) or something |
| 16:30 | duncanm | raek: oh? |
| 16:30 | duncanm | nice! |
| 16:30 | duncanm | raek: false, or true? |
| 16:31 | semperos | doh |
| 16:31 | semperos | I just found that in the source too :) |
| 16:31 | semperos | it's actually part of leiningen's project.clj |
| 16:31 | semperos | checks for it here: https://github.com/technomancy/leiningen/blob/master/src/leiningen/uberjar.clj#L74 |
| 16:31 | raek | https://github.com/technomancy/leiningen/blob/master/sample.project.clj |
| 16:32 | raek | I think this file contains all leiningen core options |
| 16:34 | duncanm | doesn't seem to work, that option |
| 16:34 | duncanm | lein deps still clears the directory |
| 16:35 | raek | duncanm: which lein version are you using? |
| 16:35 | duncanm | Leiningen 1.4.2 on Java 1.6.0_24 Java HotSpot(TM) 64-Bit Server VM |
| 16:35 | carllerche | can leiningen compile java classes as well (If I want a very simple java class?) |
| 16:36 | carllerche | oops, i should have RTFM closer :( sorry |
| 16:36 | duncanm | raek: i upgraded to 1.5.2 |
| 16:36 | raek | duncanm: current stable version is 1.5.2. maybe this was added recently |
| 16:37 | duncanm | okay |
| 16:37 | duncanm | raek: yeah, upgrading worked |
| 16:37 | duncanm | thanks |
| 16:37 | raek | carllerche: yes, if you set :java-source-path "src" in your project.clj, .java files in src/ will be compiled too |
| 16:38 | raek | you can trigger the compilation with "lein javac", I think |
| 16:38 | carllerche | thanks |
| 16:38 | semperos | duncanm: did you set it to true or false? |
| 16:38 | raek | (this happens automatically when you do lein jar) |
| 16:38 | duncanm | semperos: true |
| 16:38 | semperos | k |
| 16:39 | raek | oh, I accidentally inverted that one... sorry! :) |
| 16:41 | semperos | no worries, I didn't see that duncanm needed to upgrade his leiningen version |
| 16:41 | semperos | before I asked that question |
| 16:43 | devn | fliebel: link me to genshi? |
| 16:43 | fliebel | devn: Not Clojure: http://genshi.edgewall.org/ |
| 16:44 | fliebel | devn: Basically, the template language is written with xml attributes. |
| 16:50 | duncanm | is there a way to declare required arguments using clojure.contrib.command-line? |
| 16:55 | S11001001 | duncanm: make the default expr throw? |
| 17:08 | devn | fliebel: im sort of inclined to continue with my ridiculous idea i had awhile back |
| 17:08 | fliebel | devn: Which one? |
| 17:09 | devn | fliebel: the idea of making each template its own ns |
| 17:09 | devn | or dynamically name a template based on its namespace or something like that |
| 17:09 | devn | i have code...somewhere |
| 17:09 | fliebel | devn: This is with the closure thing? |
| 17:09 | devn | nah, that's different |
| 17:10 | devn | Google Closure templates are a good fit for clojure because they're composable |
| 17:10 | devn | or...they will be |
| 17:10 | devn | when we're done with them |
| 17:11 | fliebel | that sound cool, although I still don't like the mini language these templates invent. |
| 17:11 | fliebel | devn: So what's with the namespaces? You want to (use 'template)? |
| 17:11 | devn | fliebel: im trying to remember now! :) |
| 17:12 | devn | fliebel: the idea was you write some piece collection of template pieces in different namespaces |
| 17:13 | devn | err some collection of template pieces in a namespace |
| 17:13 | devn | so doctype, etc. falls under a namespace, and then you would mix in at a lower namespace |
| 17:13 | xian | Hi, I installed slime, slime-repl and clojure-mode via ELPA (I am using the Emacs Starter Kit) on Emacs 23.2.1 and installed swank-clojure via leiningen. Slime and swank seem to work generally, but when executing C-c C-c (slime-compile-defun), I get a "no such namespace" error (details: http://pastebin.com/brsZ0pns). |
| 17:14 | hugod | xian, compile the file (C-c C-k) at least once to define the namespace |
| 17:15 | devn | or C-x C-e at the end of the (ns)| macro |
| 17:15 | xian | hugod: Compiling the file works, but unfortunately I get the same error again when calling slime-compile-defun. |
| 17:16 | devn | fliebel: really all i want it haml but with the ability to use clojure inside of it |
| 17:16 | devn | s/it/is |
| 17:16 | sexpbot | <devn> fliebel: really all i want is haml but wish the abilisy to use clojure inside of is |
| 17:16 | fliebel | devn: Sounds fun. I need to sleep now though. See you later. You might want to check these: http://thinkslate.com/ http://code.google.com/p/molehil/ https://github.com/nakkaya/static |
| 17:16 | hugod | xian, then slime has failed to parse the namespace declaration |
| 17:17 | devn | fliebel: why thinkslate? :) |
| 17:17 | fliebel | devn: Because that seems to be your website, and it also seems to be down. |
| 17:17 | devn | :D |
| 17:17 | xian | hugod: I'm probably too much of a clojure noob, but I don't even have an explicit namespace declaration in my source file, just a single defun. |
| 17:17 | devn | fliebel: im migrating that domain to google apps |
| 17:18 | devn | fliebel: but yes you're quite right |
| 17:18 | xian | Oh I mean defn of course, sorry. |
| 17:18 | devn | i need to get my damned websites up |
| 17:18 | devn | xian: then you should be able to compile that into the user namespace |
| 17:18 | fliebel | devn: Me too, that is why I'm doing bakers again :) |
| 17:18 | devn | fliebel: :) |
| 17:18 | devn | gotta run, talk to you soon fliebel |
| 17:18 | fliebel | bye |
| 17:19 | hugod | xian, you need a (ns ...) form for C-c C-c to work |
| 17:21 | raek | xian: I think C-M-x work even without an ns form. it evaluates the topmost form that contains the point. |
| 17:21 | xian | Ah all right. I just added (ns user) and it magically worked. Thank you both very much, hugod and devn, for helping a noob out. |
| 17:21 | raek | ...but you should of course look into namespaces sooner or later |
| 17:22 | raek | this is a great guide: http://blog.8thlight.com/articles/2010/12/6/clojure-libs-and-namespaces-require-use-import-and-ns |
| 17:22 | xian | raek: Yeah, I found out about that too, but I just like to use C-c C-c (it's quite easy to reach). |
| 17:23 | xian | I'm going to look into that, thanks. |
| 19:29 | miwillhite | I'm brand new to clojure and trying to understand this: |
| 19:29 | miwillhite | (java.io.File. logs-path) |
| 19:29 | miwillhite | whats with that dot after File |
| 19:30 | brehaut | miwillhite: it treats it as a constructor |
| 19:30 | brehaut | otherwise java.io.File is a reference to the class |
| 19:30 | miwillhite | so thats like instantiating a new File object? |
| 19:30 | brehaut | thats exactly what it is |
| 19:30 | miwillhite | and the dot triggers that? |
| 19:30 | brehaut | yes |
| 19:30 | miwillhite | great, thank you |
| 19:30 | miwillhite | is that a clojurism? |
| 19:31 | brehaut | yes |
| 19:31 | brehaut | (java.net.URL. "http://example.com") |
| 19:31 | brehaut | ,(java.net.URL. "http://example.com/") |
| 19:31 | clojurebot | #<URL http://example.com/> |
| 19:31 | miwillhite | ah wow, cool |
| 19:31 | brehaut | theres also an older form you might enocunter in the wild |
| 19:31 | brehaut | ,(new java.net.URL "http://example.com/") |
| 19:31 | clojurebot | #<URL http://example.com/> |
| 19:31 | miwillhite | the comma |
| 19:32 | miwillhite | ? |
| 19:32 | miwillhite | oh the 'new' |
| 19:32 | miwillhite | nm |
| 19:32 | miwillhite | whats the comma do? |
| 19:32 | brehaut | the comma tells clojurebot to eval the sexp |
| 19:32 | miwillhite | oh haha |
| 19:33 | miwillhite | well thanks a lot |
| 19:33 | miwillhite | I've found this room to be very helpful when I have questions |
| 19:33 | brehaut | likewise |
| 19:34 | pdk | yeah |
| 19:34 | pdk | we've found you very helpful too like brehaut said :) |
| 19:35 | brehaut | haha |
| 20:00 | lawfulfalafel | I am trying to test code in the clojure repl and I keep getting classpath errors |
| 20:00 | lawfulfalafel | such as: |
| 20:00 | lawfulfalafel | java.io.FileNotFoundException: Could not locate clojure/contrib/duck_streams__init.class or clojure/contrib/duck_streams.clj on classpath: (NO_SOURCE_FILE:0) |
| 20:01 | lawfulfalafel | I've tried creating a .clojure file that points to /usr/share/java/clojure-contrib.jar |
| 20:01 | amalloy | $google getting started with clojure raynes |
| 20:01 | sexpbot | First out of 72 results is: An indirect guide to getting started with Clojure » Bathroom ... |
| 20:01 | sexpbot | http://blog.raynes.me/%3Fp%3D48 |
| 20:01 | amalloy | argh. i always forget that url is broken |
| 20:01 | amalloy | http://blog.raynes.me/?p=48 |
| 20:02 | amalloy | lawfulfalafel: long story short, use lein or cake. even for teeny tiny one-file "projects", having a build tool makes your life easier |
| 20:03 | miwillhite | Another question if you don't mind…can someone explain the [x# ~x] in the snippet below?: |
| 20:03 | miwillhite | (defmacro dbg[x] `(let [x# ~x] (println "dbg:" '~x "=" x#) x#)) |
| 20:03 | miwillhite | I don't understand what the # does |
| 20:03 | miwillhite | or the ~ |
| 20:04 | Somelauw | # makes an anynomous function with implicit parameter % |
| 20:04 | miwillhite | I've read that the ~ "unquotes", but I'm not sure I understand that either |
| 20:04 | lawfulfalafel | amalloy: really? I have never used one of those before, so it's hard for me to imagine the benefits |
| 20:04 | lawfulfalafel | which one would you reccomend? |
| 20:05 | Somelauw | But I don't understand how it is used there either. |
| 20:06 | miwillhite | Somelauw, I've seen #() as an anon function |
| 20:06 | Somelauw | I think it is just a regular variable. |
| 20:06 | miwillhite | but never after a variable |
| 20:06 | Somelauw | with a symbol in its name |
| 20:06 | Somelauw | so x# is the variable name |
| 20:06 | miwillhite | ah, I think you are right |
| 20:06 | miwillhite | thats weird |
| 20:07 | technomancy | ,`x# |
| 20:07 | clojurebot | x__1650__auto__ |
| 20:07 | technomancy | generates a unique symbol to prevent accidental variable capture in a macro |
| 20:10 | amalloy | i've used cake for quite a while but i think i have to recommend lein |
| 20:11 | miwillhite | user=> `(1 2 (list 3 4)) |
| 20:11 | miwillhite | (1 2 (clojure.core/list 3 4)) |
| 20:11 | miwillhite | user=> `(1 2 ~(list 3 4)) |
| 20:11 | miwillhite | (1 2 (3 4)) |
| 20:11 | miwillhite | user=> `(1 2 ~@(list 3 4)) |
| 20:11 | miwillhite | (1 2 3 4) |
| 20:11 | miwillhite | that makes sense |
| 20:11 | miwillhite | looking at unquote |
| 20:36 | lawfulfalafel | sorry, but how does one add the lein self-install bin to bash? |
| 20:37 | lawfulfalafel | I am trying to be able to do the "lein new blah" command |
| 20:37 | technomancy | lawfulfalafel: just put it anywhere on your $PATH. /usr/local/bin/lein if you like |
| 20:40 | lawfulfalafel | technomancy: thank you |
| 20:41 | technomancy | np |
| 20:52 | joshua__ | Ouch. It seems I don't have a summer internship =/ I'm a little surprised I didn't even get a phone interview anywhere. |
| 21:41 | amalloy | joshua__: come work on 4clojure! nobody will pay you a cent but you can tell people you worked on 4clojure |
| 23:24 | bhenry | is anybody using this? https://github.com/overtone/live-coding-emacs |
| 23:26 | bhenry | there is a rather annoying "feature" that shows lambdas and f(expr) notations. http://i.imgur.com/MvOHU.png |
| 23:26 | bhenry | i'd like to know where in .emacs.d i have to go to turn it off. |
| 23:32 | tomoj | grep -r λ . |
| 23:44 | bhenry | tomoj thank you. i couldn't figure out how to search the contents of .el files. |
| 23:56 | amalloy | find -name '*.el' | xargs grep whatever |
| 23:56 | amalloy | would be a little more fine-grained |
