00:01 Demosthenes: second test uses the key from the text as the symbol key for the hash
00:01 yoklov: yeah i figured that out
00:01 yoklov: i dont think my method will be any more efficient, but it might give you a different way to think about it or something, who knows.
00:01 yoklov: sec.
00:01 Demosthenes: i appreciate the review, this has been eating at me ;]
00:02 Demosthenes: i was thinking of a recursive parser... processing a list of directives like '((a "regexp")(b "regexp")(c "regexp))
00:02 yoklov: no problem, i should be studying for a differential equations exam lol
00:02 yoklov: yeah
00:02 yoklov: okay
00:02 yoklov: that actually helps me out
00:02 Demosthenes: if you get to the end of the directives successfully, keep the record
00:02 yoklov: basically the way i was doing it was
00:03 yoklov: a list of regexps, and a tail recursive function that if the list is null
00:03 Demosthenes: but then by specifying that, i got back to (match ....)
00:03 yoklov: return a hash
00:03 yoklov: if uh, they match
00:03 Demosthenes: kinda same line of thought there
00:03 yoklov: then set the hash, and call the next part of the function
00:03 yoklov: with cdr list cdr regexp list
00:03 Demosthenes: i just had the impression match might save writing a function using an existing feature
00:04 yoklov: honestly i have no idea what match does.
00:04 yoklov: hold on i'll look
00:04 Demosthenes: look at the doc.
00:04 Demosthenes: its slick... if my data were lists... i think it could match it all
00:05 yoklov: hm
00:05 yoklov: oh wow
00:05 yoklov: thats pretty cool.
00:06 yoklov: its like haskell pattern matching
00:06 yoklov: possibly older who knows
00:06 yoklov: though
00:06 yoklov: i dont really know if its what you want.
00:06 Demosthenes: did you look at TXR? ;]
00:07 yoklov: clearly not
00:07 yoklov: looking now
00:07 Demosthenes: somewhere i think there's a happy middle, between recursion, named capture (missing in racket but in perl), txr, and match
00:07 (quit) tauntaun: Quit: Ex-Chat
00:08 yoklov: i guess to be honest, i dont understand any of this.
00:08 yoklov: it seems fairly slick though.
00:09 yoklov: either way dude
00:09 yoklov: what you want to do doesnt seem that hard with recursion
00:10 yoklov: wait i didnt go far enough in txr
00:16 Demosthenes: its basically what txr summarized, reversing printf ;]
00:17 Demosthenes: imagine being given business records in text format from an app you don't control
00:17 Demosthenes: its like reading a report back into rows...
00:17 yoklov: here, basically this was my idea http://pastebin.com/16yiwwQ0 . you'd have to design the noise and out of order functions. noise should porbably be a regexp or something that checks if a string has a particular thing that you know makes it not noise, out of order you could define by having some other ugly mutable variable in your function
00:17 yoklov: or by having a 'seen' variable be passed along with iterate
00:17 yoklov: and maybe pass that to out-of-order?
00:18 Demosthenes: i think the recursion is the answer
00:18 Demosthenes: also lets me do things like "lookahead"
00:19 yoklov: i mean, if you do non tail-call recursion you'll have a significant performance hit
00:19 Demosthenes: thing is, perl's regexp engine you can put in debug mode, and it'll print verbatim the processing of a regexp, but racket doesn't seem to do that anywhere.... so if a regexp fails its a mystery
00:19 yoklov: especially if your text to parse is long enouhg
00:19 yoklov: it returns false i thought
00:19 Demosthenes: sure, but in a long complex regexp, how do you know where it failed?
00:19 yoklov: you could define somethign like
00:19 yoklov: oh
00:19 yoklov: i see.
00:20 Demosthenes: one of the reasons i'm looking for false positives or failures is to detect new record types when they change their code....
00:20 yoklov: well, hm
00:20 yoklov: i mean.
00:20 yoklov: there's a solution to this problem
00:20 yoklov: and i think that solution may be macros.
00:20 Demosthenes: it may be
00:21 Demosthenes: but i think recursion and a directive list is a great start
00:21 Demosthenes: really, its designing a general purpose event driven parser
00:21 yoklov: i see.
00:21 Demosthenes: not that i was thinking that when i started :P
00:21 yoklov: lol
00:21 Demosthenes: it was more of "oh, this is text. i ought to be able to parse it out easy"
00:22 Demosthenes: anyhow, afk for a bit
00:22 yoklov: yeah i'm out too
00:22 Demosthenes: thanks for the input and perspecive
00:22 yoklov: no problem
00:22 yoklov: thanks for the distraction
00:22 yoklov: catch you later.
01:22 (join) realitygrill_
01:25 (quit) realitygrill: Ping timeout: 250 seconds
01:25 (nick) realitygrill_ -> realitygrill
01:41 (quit) yoklov: Quit: Leaving.
01:54 (quit) geoffhill: Quit: geoffhill
02:49 (part) cpach
03:08 (quit) ckrailo: Quit: Computer has gone to sleep.
03:19 (quit) em: Ping timeout: 260 seconds
04:02 (quit) realitygrill: Quit: realitygrill
05:18 (join) PLT_Notify
05:18 PLT_Notify: racket: master Mike Sperber * d582dd5 (2 files in 2 dirs): Fix bug in vector signatures. - http://bit.ly/mUj8jx
05:18 (part) PLT_Notify
06:00 (join) MayDaniel
06:01 (join) dzhus
06:09 (quit) MayDaniel: Read error: Connection reset by peer
06:16 (quit) dzhus: Read error: Operation timed out
06:46 (join) lucian
07:03 (join) mceier
07:26 (join) lucian_
07:29 (quit) lucian: Ping timeout: 264 seconds
07:38 (join) masm
07:40 (join) yoklov
07:45 (nick) lucian_ -> lucian
07:46 (quit) Demosthenes: Quit: leaving
08:47 (quit) lucian: Ping timeout: 276 seconds
08:50 (join) lucian
08:50 (join) tauntaun
08:52 (quit) tauntaun: Client Quit
09:08 (quit) Triplefault: Quit: Leaving
09:08 (join) MayDaniel
09:16 (quit) yoklov: Quit: Leaving.
09:35 (quit) hyko: Read error: Connection reset by peer
09:35 (join) hyko
09:57 (join) mithos28
10:05 (quit) mithos28: Quit: mithos28
10:10 (join) yoklov
10:34 (join) anRch
10:35 (join) tauntaun
11:01 (quit) DT``: Ping timeout: 250 seconds
11:12 (quit) anRch: Quit: anRch
11:14 (join) DT``
11:15 (quit) dnolen: Quit: dnolen
11:27 (join) emma
11:32 (nick) emma -> em
12:04 (quit) MayDaniel: Read error: Connection reset by peer
12:19 (join) dnolen
12:20 (join) realitygrill
12:59 (join) jonrafkind
13:35 (join) ckrailo
14:11 (quit) qha: Ping timeout: 260 seconds
14:16 (join) qha
14:53 (quit) realitygrill: Ping timeout: 240 seconds
15:19 (join) realitygrill
15:34 (join) PLT_Notify
15:34 PLT_Notify: racket: master Robby Findler * 551c686 (1 files in 1 dirs): add in the supported ->i forms
15:34 PLT_Notify: racket: master Robby Findler * d6ef130 (13 files in 3 dirs): cleaned up framework docs
15:34 PLT_Notify: racket: master commits d582dd5...d6ef130 - http://bit.ly/jnRZlU
15:34 (part) PLT_Notify
15:42 (join) MayDaniel
15:46 (quit) yoklov: Quit: Leaving.
15:50 (quit) MayDaniel: Read error: Connection reset by peer
15:59 (join) yoklov
16:17 (quit) chemuduguntar: Remote host closed the connection
16:51 (join) lucian_
16:51 (quit) lucian: Read error: Operation timed out
17:08 (nick) lucian_ -> lucian
17:14 (quit) yoklov: Quit: Leaving.
17:22 (quit) rekahsoft: Ping timeout: 240 seconds
17:24 (join) rekahsoft
17:28 (quit) rekahsoft: Ping timeout: 240 seconds
17:34 (quit) mceier: Quit: leaving
17:48 (quit) realitygrill: Quit: realitygrill
17:55 (quit) masm: Quit: Leaving.
17:58 (join) masm
18:07 (quit) masm: Quit: Leaving.
18:08 (join) masm
18:09 (join) yoklov
18:16 (quit) DT``: Ping timeout: 250 seconds
18:30 (join) DT``
19:04 (join) chemuduguntar
19:16 (join) mithos28
19:24 (quit) ckrailo: Quit: Computer has gone to sleep.
19:28 (join) PLT_Notify
19:28 PLT_Notify: racket: master Robby Findler * 2bba4e1 (6 files in 2 dirs): fixed up the graphics/ collection's docs
19:28 PLT_Notify: racket: master Robby Findler * bc9c44a (2 files in 1 dirs): graphics/turtles: fixed up for an 'xor-less world ...
19:28 PLT_Notify: racket: master commits d6ef130...bc9c44a - http://bit.ly/kn5hvk
19:28 (part) PLT_Notify
19:45 (join) lucian_
19:48 (quit) lucian: Ping timeout: 260 seconds
19:53 (join) rpr
19:56 rpr: Ok Typed Racket is just the best thing I've seen in a loooong while. O U T S T A N D I N G
19:58 (nick) lucian_ -> lucian
20:00 yoklov: really?
20:00 yoklov: i don't 'get' it
20:01 yoklov: it seemed like a bit of a hassle.
20:01 jonrafkind: types are a hassle?
20:01 yoklov: well, i mean, to someone who doesn't know about them, they seem that way
20:02 jonrafkind: I suppose, you've never used a type language before?
20:03 yoklov: some C, but i dont think thats what you men
20:03 yoklov: mean*
20:04 yoklov: oh, and the first two chapters of a Haskell book
20:04 jonrafkind: yea any complex program in C basically abuse the type system
20:04 yoklov: that talked a ton about how awesome types were but didnt get to the point
20:04 jonrafkind: you should try writing some programs in haskell or ocaml
20:05 jonrafkind: if you can get the program to type check there are lots of benefits
20:05 jonrafkind: of course, if you have a type error sometimes it seems like a waste of time trying to fix it
20:05 yoklov: yeah, thats the problem i usually had
20:05 yoklov: in typed racked
20:06 jonrafkind: yes perhaps typed racket is not mature enough yet to be useable by more people, i dont really know
20:06 yoklov: racket*.  all my programs had a whole mess of bugs
20:06 yoklov: which is probably personal more than the maturity of typed racket
20:09 (join) rekahsoft
20:14 (join) geoffhill
20:17 yoklov: hm, i had wondered if monads were a thing that you could implement in racket
20:18 yoklov: says they're type classes on wikipedia so it would seem like that could be doable in typed racket
20:18 yoklov: not that i actually understand monads :/
20:19 (join) ckrailo
20:23 (quit) lucian: Ping timeout: 240 seconds
20:25 (quit) mithos28: Quit: mithos28
20:25 (join) lucian
20:26 rpr: There is something special happening in Typed Racket that I don't have enough horsepower to put my finger on.  But the type system is pretty damn sophisticated.  Reminds me of Scala's. AND they seamless merge in runtime contracts into typed racket.  Its all far more than I expected.
20:30 rpr: FWIW today I've been converting all my code over to typed/racket.  Going ok in general.
20:30 yoklov: have you found that typed racket takes longer to load up initially?  that was another reason that i was turned off by it
20:31 yoklov: i'm just wondering if that's just my system or whatever
20:35 rpr: Yes a somewhat slower.  I use emacs + geiser and its fast enough for me.  I have a decent amount of code.
20:38 (quit) masm: Ping timeout: 240 seconds
20:40 (join) realitygrill
20:40 rpr: Once loaded, while converting a module of developing, once the typer is up to speed I've found type checking to be good enough.
20:46 bremner_: jonrafkind: fyi 5.1.1 uploaded to debian
20:48 yoklov: yeah, the reason I wanted to write in typed racket in the first place was that something in the docs had mentioned typed racket having better performance, and I was unsatisfied with the performance of my mandelbrot set drawer lol
20:49 yoklov: the slowdown on loading it initially was enough to make me not end up figuring it out though.
21:12 (join) mithos28
21:18 (quit) tauntaun: Quit: Ex-Chat
22:29 (join) wwwd
22:46 (quit) mithos28: Quit: mithos28
23:05 (quit) lucian: Remote host closed the connection
23:05 (quit) wwwd: 
23:27 (quit) ckrailo: Quit: Computer has gone to sleep.
23:30 (quit) rekahsoft: Ping timeout: 276 seconds
23:30 (join) mithos28
23:35 (quit) yoklov: Quit: Leaving.
23:41 (join) ckrailo
23:41 (quit) rpr: Remote host closed the connection
23:51 (join) rekahsoft