May 2014 - Beginners - Haskell.org

Data.Stream interleave implementation question
by Bryan Brady 19 Aug '14

19 Aug '14

I'm in the process of learning Haskell and am implementing my own Stream type. ( I'm working my way through homework problems I found at http://www.seas.upenn.edu/~cis194/hw/06-laziness.pdf ) One of the questions is: "Deﬁne the stream ruler :: Stream Integer which corresponds to the ruler function 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, 4, . . . where the nth element in the stream (assuming the ﬁrst element corresponds to n = 1) is the largest power of 2 which evenly divides n. " I got pretty close, but ended up looking at Data.Stream to nail down my mistake. My original solution was: ----- data Stream a = Cons a (Stream a) deriving (Eq, Ord) streamRepeat :: a -> Stream a streamRepeat a = Cons a (streamRepeat a) interleaveStreams :: Stream a -> Stream a -> Stream a interleaveStreams (Cons a as) (Cons b bs) = Cons a (Cons b (interleaveStreams as bs)) ruler :: Stream Integer ruler = foldr1 interleaveStreams (map streamRepeat [0..]) ----- Attempting to evaluate 'ruler' in ghci does not finish. I knew something was wrong with my implementation of interleaveStreams because it worked when I expanded it out manually for a few steps. The way Data.Stream implements interleave is: interleaveStreams (Cons a as) bs = Cons a (interleaveStreams bs as) I don't understand why this works and why my original implementation doesn't. I figure if the latter works, the former should as well. Here is my reasoning: In the latter definition, Cons a (interleaveStreams bs as), (interleaveStreams bs as) is a thunk. The thunk should only be evaluated when it is needed. In my original definition, (interleaveStreams as bs) is a thunk. The difference is an extra Cons (e.g., Cons b). It seems like the additional Cons is causing problems, I just don't understand why. Can anyone point out what I'm missing? thanks, bb

7 10

Functors in Haskell
by Dimitri DeFigueiredo 07 Jul '14

07 Jul '14

Hi All, I'm trying to merge in my head the concepts of: - functors from category theory; with, - the Functor typeclass in haskell. Here's what I have come up with: A Functor F in haskell is a type constructor that provides a functor map (fmap) and that maps Hask (the category of haskell types) into a category generated by the Functor. For example, Maybe maps the category of all haskell types "a" (i.e. Hask) into the category of all types of the form "Maybe a" by providing an implementation of fmap. Neglecting the fact that apparently Hask isn't a category, does this sound reasonable? I think the insight for me here is that functors in haskell are type constructors mapping Hask into another category. Is that a good way to think about it? Thanks! Dimitri

6 12

Project euler question
by martin 06 Jun '14

06 Jun '14

Hello all, I tried to solve Problem 24 (https://projecteuler.net/problem=24) and came up with the following solution: import Data.List.Ordered import Data.Char elems = [0,1,2,3,4,5,6,7,8,9] :: [Int] x = do a <- elems b <- elems `without` [a] c <- elems `without` [a,b] d <- elems `without` [a,b,c] e <- elems `without` [a,b,c,d] f <- elems `without` [a,b,c,d,e] g <- elems `without` [a,b,c,d,e,f] h <- elems `without` [a,b,c,d,e,f,g] i <- elems `without` [a,b,c,d,e,f,g,h] j <- elems `without` [a,b,c,d,e,f,g,h,i] return [a,b,c,d,e,f,g,h,i,j] without a b = minus a ( sort b) solution = filter isDigit $ show $ (x !! 1000001) -- "2783915640" PE tells me that this is wrong, and I peeked the correct answer, which is 2783915460 (the 4 and 6 are swapped). So I tried to find out where the correct answer is in my list x and added y = filter (\(x,y) -> x == "2783915460") $ zip (map (filter isDigit . show) x) [1..] -- [("2783915460",1000000)] How can that be? "solution" tells me that the millionth element is "2783915640" but "y" tells me that "2783915460" is at the millionth position? I just cannot see it.

8 15

Static records (CSV tables as modules)
by Mark Fredrickson 03 Jun '14

03 Jun '14

Hello, I writing a program that operates on some static data, currently saved in CSV files. Right now I parse the files at run time and then generate hashmap tables to connect the different data. Since I'm only ever operating on static data, I'm wondering if I can generate module files that capture the records as a sum type. To access the fields of the records, I could then imagine functions that exhaustively map the constructors to the data. Do any tools to generate .hs files from CSV or other formats exist? Insofar as this question has been asked before, the recommendation is "use Template Haskell", which makes sense, but is a less complete solution than I'm hoping for. Here's example of what what would be generated in literal Haskell source: -- input.csv name, age, someValue "abc", 1, 3.0 "xyz3", 99, -5.9 -- Input.hs module Input where data Row = R1 | R2 name :: Row -> String name R1 = "abc" name R2 = "xyz3" age :: Row -> Integer age R1 = 1 age R2 = 99 -- likewise for the function someValue Thanks, -M

3 3

installing cabal 1.20 on linux mint debian edition (LMDE)
by Ovidiu Deac 31 May '14

31 May '14

I'm trying to install cabal 1.20 like this: I installed ghc and ghc-prof from apt repository so I got the following: ii ghc 7.6.3-5 amd64 The Glasgow Haskell Compilation s ii ghc-prof 7.6.3-5 amd64 Profiling libraries for the Glasg Then I git cloned cabal and checked out Cabal-v1.20.0.0 Then I ran bootstrap.sh in cabal-install dir and I get the following error: Using gcc for C compiler. If this is not what you want, set CC. Using /usr/lib/gcc/x86_64-linux-gnu/4.8/collect2 instead. Checking installed packages for ghc-7.6.3... deepseq is already installed and the version is ok. time is already installed and the version is ok. Cabal is already installed and the version is ok. transformers is already installed and the version is ok. mtl is already installed and the version is ok. text is already installed and the version is ok. parsec is already installed and the version is ok. network is already installed and the version is ok. HTTP is already installed and the version is ok. zlib is already installed and the version is ok. random is already installed and the version is ok. stm is already installed and the version is ok. [...skipping many lines] package cabal-install-1.20.0.0 requires network-2.4.1.2 Building cabal-install-1.20.0.0... Preprocessing executable 'cabal' for cabal-install-1.20.0.0... <command line>: cannot satisfy -package-id network-2.4.1.2-fc99093587d92370c7febe034504fb40: network-2.4.1.2-fc99093587d92370c7febe034504fb40 is shadowed by package network-2.4.1.2-040cee5ece44014a8574cb3f87b1eec4 (use -v for more information) Error during cabal-install bootstrap: Building the cabal-install package failed. What am I doing wrong here? Thanks, Ovidiu

5 8

Best way to have a fresh&updated Haskell installation
by Christian Espinoza L. 31 May '14

31 May '14

Hi guys, like subject says , how can I get it? I'm using Debian Linux, and I need a guide to have a updated haskell platform, easy to update when will necessary. Haskell Platform is totally out of date, I know is required compile all, but I need a guide in order to get a standard way. Thanks in advance. Christian.

2 1

Using "cabal test" and getting "cabal: Prelude.read: no parse"
by Daniel King 30 May '14

30 May '14

Hi all, I created a cabal project with the attached cabal file. I run: cabal configure --enable-tests cabal build cabal test and then I get the error: cabal: Prelude.read: no parse I'm not sure how to debug this any further. If I execute: ./dist/build/scientific-pl-testsStub/scientific-pl-testsStub with any input I've tried (numbers, array notation, string notation, etc.) and I get that same error. I can post the whole project if that is helpful. Thanks for the debugging tips! The full log is: danking@spock # cabal configure --enable-tests Resolving dependencies... Configuring scientific-pl-0.1.0.0... danking@spock # cabal build Building scientific-pl-0.1.0.0... Preprocessing test suite 'scientific-pl-tests' for scientific-pl-0.1.0.0... [1 of 3] Compiling SPLData ( SPLData.hs, dist/build/SPLData.o ) [2 of 3] Compiling SPLEval ( SPLEval.hs, dist/build/SPLEval.o ) [3 of 3] Compiling Tests ( tests/Tests.hs, dist/build/Tests.o ) In-place registering scientific-pl-tests-0.1.0.0... [1 of 1] Compiling Main ( dist/build/scientific-pl-testsStub/scientific-pl-testsStub-tmp/scientific-pl-testsStub.hs, dist/build/scientific-pl-testsStub/scientific-pl-testsStub-tmp/Main.o ) Linking dist/build/scientific-pl-testsStub/scientific-pl-testsStub ... Preprocessing executable 'scientific-pl' for scientific-pl-0.1.0.0... [1 of 3] Compiling SPLData ( SPLData.hs, dist/build/scientific-pl/scientific-pl-tmp/SPLData.o ) [2 of 3] Compiling SPLEval ( SPLEval.hs, dist/build/scientific-pl/scientific-pl-tmp/SPLEval.o ) [3 of 3] Compiling Main ( Main.hs, dist/build/scientific-pl/scientific-pl-tmp/Main.o ) Linking dist/build/scientific-pl/scientific-pl ... danking@spock # cabal test Running 1 test suites... Test suite scientific-pl-tests: RUNNING... cabal: Prelude.read: no parse 1 danking@spock # -- Dan King

5 8

compilation joys
by Elise Huard 30 May '14

30 May '14

Hi, I've been compiling this project https://github.com/mlesniak/game.git It uses Hipmunk (2D collision detection/physics framework) and OpenGL. Now the problem is that Hipmunk uses Data.StateVar, and OpenGL/GLUT use Graphics.Rendering.OpenGL.GL.StateVar Both have pretty much the same code and functionality, but using both together gives problems with conflicting types. My current solution was to locally fork Hipmunk and make it use Graphics.Rendering.OpenGL.GL.StateVar (because the other way round - making OpenGL use Data.StateVar, is a cascading nightmare, lots of libs to patch), and this works. My question: is this the only possible solution? Is there some way that doesn't involve patching one library or the other? The other thing I had to locally patch is to force cabal to use freeglut instead of the local Mac OS X GLUT. I did this by again, locally forking GLUT (long live cabal sandbox add-source!) and adding extra-lib-dirs: /usr/local/Cellar/freeglut/2.8.1/lib which is the local homebrew freeglut dir. Is there a better, more reproducible way to do this? Thank you, Elise

2 1

Network.HTTP.Conduit - body in REST/POST request
by Miro Karpis 29 May '14

29 May '14

Hi, I'm having difficulties with sending body parameters/values in a POST request. It seems that the server does not receive in body. ideas/comments very welcome ;-) cheers, m. {-# LANGUAGE OverloadedStrings #-} import Network.HTTP.Conduit import Data.Word(Word8) import Data.ByteString.Lazy(pack) import qualified Data.ByteString.Char8 as B x :: B.ByteString x = "?value=10" post = do r <- parseUrl "http://postcatcher.in/catchers/5385d4e0b6887c0200000071" putStrLn $show r let request = r { secure = True , method = "POST" , requestBody = RequestBodyBS s , requestHeaders = (requestHeaders r) ++ [("Content-Type", "application/json")]} putStrLn $ show request response <- withManager $ httpLbs request print $ responseBody response

2 2

Re: [Haskell-beginners] apply a function to all params of another function
by Brent Yorgey 28 May '14

28 May '14

On Mon, May 26, 2014 at 04:46:29PM +0200, Kees Bleijenberg wrote: > In have a lot of functions in a program with params that are alle Int's. > f :: Int -> Int -> Int > g :: Int -> Int -> Int > h :: Int -> Int -> Int -> Int > ... > and convertParam :: Int -> Int -- say (+1) > I want to call functions like f, g and h, but apply convertParam to all > params first. > f2 a b c = f (convertParam a) (convertParam b) (convertParam c) > > I tried: > run1p f conv = \a -> f (conv a) -- for functions with one > param > run2p f conv = \a b -> f (conv a) (conv b) -- for functions with two > params > . > Can you do this in a more generalized way (using currying?) > Any ideas? class Convertible t where convert :: t -> t instance Convertible Int where ... instance Convertible t => Convertible (Int -> t) where ... I'll let you fill in the ... ! =) Note this only works well because the base case is Int. You could also add some extra base cases for other concrete return types. However, it is quite difficult to give a Convertible instance which applies to "all types which are not functions", so this will only work if you are willing to make one instance for each concrete result type your functions use. -Brent

2 1