October 2009 - Haskell-Cafe

Re: ghci can't find Paths_ module created by cabal
by Thomas Hartman 26 Oct '09

26 Oct '09

from my bash hints file: thartman@ubuntu:~/haskellInstalls/gitit>thartman_ghci_with_data_files Loading a package with data files in ghci See http://neilmitchell.blogspot.com/2008/02/adding-data-files-using-cabal.html Snip: The above method works well after a program has been installed, but is harder to work with while developing a program. To alleviate these problems, we can add our own Paths module to the program, for example: module Paths_hoogle where getDataFileName :: FilePath -> IO FilePath getDataFileName = return Place this module alongside all the other modules. While developing the program our hand-created Paths module will be invoked, which says the data is always in the current directory. When doing a Cabal build, Cabal will choose its custom generated Paths module over ours, and we get the benefits of Cabal managing our data. 2009/10/21 Robin Green <greenrd(a)greenrd.org>: > I'm trying to run cabal-install in ghci so I can run it under the ghci > debugger. However I try it, though, I always get one of these two > errors: > > /src/greenrd/darcs/cabal-install/Main.hs:70:17: > Could not find module `Paths_cabal_install': > Use -v to see a list of the files searched for. > Failed, modules loaded: none. > > or > > <command line>: > Could not find module `Main': > Use -v to see a list of the files searched for. > > -- > Robin > _______________________________________________ > Libraries mailing list > Libraries(a)haskell.org > http://www.haskell.org/mailman/listinfo/libraries >

2 1

How come pattern match does not recognize this code style?
by Magicloud Magiclouds 26 Oct '09

26 Oct '09

Hi, In case of this: case bala of [ bala , bala , bala ] -> bala bala bala _ -> bala ghc 6.10.4 would fail to compile, because of the wrong indent of line ", bala". -- 竹密岂妨流水过山高哪阻野云飞

3 3

ANN: GlomeVec, IcoGrid
by Jim Snow 26 Oct '09

26 Oct '09

I just uploaded two packages to hackage. The first package [1] is the vector library used by my haskell ray-tracer, Glome, which has been neglected of late. It's one of the first things I ever wrote in Haskell, so some of the code looks a little strange to me now, but it works pretty well for a wide variety of computational geometry tasks. I also included a perlin noise implementation. I released GlomeVec so that I would have a reasonable vector library to work off of for my other project, which is the IcoGrid library [2]. IcoGrid (Icosahedron Grid) is a library for dealing with grids of hexagons and pentagons wrapped around a sphere. Individual grid cells are identified by an integer, and the library can find all the neighbors of a given cell, it can find all the places where three cells intersect at a point, and it can also tell you where in 3-D space the center of the cell is. (The cells are positioned on the surface of a sphere with unit 1.) The one very useful function I haven't implemented yet is, given an arbitrary point, find the cell that the point "belongs" in. I posted a screenshot on the Haskell wiki [3] showing an OpenGL application I wrote that uses IcoGrid. My initial goal was to write some sort of simple environmental simulator that can eventually be made into a game. The grid turned out to be a pretty big project by itself, so I turned it into a stand-alone library. Both libraries have haddock docs, but they aren't showing up on hackage. Is there a trick to tell hackage it needs to generate docs, or do I just need to wait for a cron job to run? -jim [1] http://hackage.haskell.org/package/GlomeVec [2] http://hackage.haskell.org/package/IcoGrid [3] http://www.haskell.org/haskellwiki/IcoGrid

1 0

Re: What is the current state of affairs with supercompilation?
by Peter A. Jonsson 25 Oct '09

25 Oct '09

Hi Eugene, here are some answers: > I'm mainly interested in the following: > - What supercompilers do exist, except for the ones mentioned above? We have prototype implementations of supercompilers for both GHC and the Timber compiler, but they are not yet in a suitable form for public inspection. Neil Mitchell has the source code to Supero available on his homepage [1], which I think uses yhc as frontend and ghc as backend. > - What academic organizations do research in the area? Looking at where the authors of some publications on supercompilation were at the time of the publication gives some hints: - University of York: Neil Mitchell, Colin Runciman - University of Copenhagen: Morten Heine Sørensen, Robert Glück, Jens Peter Secher, Neil Jones - University of Liverpool: Alexei Lisitsa - Russian Academy of Sciences: Sergei Romanenko, Ilya Klyuchnikov - Luleå University of Technology: myself and Johan Nordlander > - What papers should be read by someone eager to learn about > supercompilation and maybe to make research contributions? If you are interested in program optimization, here's two pointers: - A Positive Supercompiler [2] - An Algorithm of Generalization in Positive Supercompilation.[3] Combine these two and you have a positive supercompiler for a first- order functional language, using the notation that is familiar to the functional programming community. More recent work on program optimization: - Neil's thesis [4] - There's a number of tutorials available from the group at University of Copenhagen, all worth a read. Alexei Lisitsa, Sergei Romanenko, and Ilya Klyuchnikov have used supercompilation to do verification: - Verification as a parameterized testing (experiments with the SCP4 supercompiler). [5] - Proving the equivalence of higher-order terms by means of supercompilation. [6] > - What are some solved problems and some open problems in > supercompilation? The algorithms are quite well-studied with respect to correctness. There's less material available on actual implementations, and how to get them to perform well. I haven't seen anything discussing code explosion and similar practical issues. I don't think anyone has mechanically verified the correctness of supercompilation. This is by no means an exhaustive list of everything, but it's something to start with. I hope it helps. Kind Regards, Peter [1] http://community.haskell.org/~ndm/supero/ [2] ftp://ftp.diku.dk/pub/diku/semantics/papers/D-300.ps.gz [3] http://citeseerx.ist.psu.edu/viewdoc/download? doi=10.1.1.49.1869&rep=rep1&type=pdf [4] http://community.haskell.org/~ndm/thesis/ [5] http://www.springerlink.com/content/c8623847257w22j9/ [6] http://pat.keldysh.ru/~roman/doc/Klyuchnikov,Romanenko-2009-- Proving.the.Equivalence.of.Higher- Order.Terms.by.Means.of.Supercompilation.pdf

2 1

unix pkg on cygwin
by Ben Fitzgerald 25 Oct '09

25 Oct '09

Hi just starting out on haskell so thought I'd try and install a few packages. from unix-2.3.2.0: $ runhaskell Setup.hs configure checking target system type... i686-pc-cygwin configure: creating ./config.status config.status: creating unix.buildinfo config.status: creating include/HsUnixConfig.h Setup.hs: Missing dependency on a foreign library: * Missing header file: HsUnix.h * Missing C library: dl This problem can usually be solved by installing the system package that provides this library (you may need the "-dev" version). If the library is already installed but in a non-standard location then you can use the flags --extra-include-dirs= and --extra-lib-dirs= to specify where it is. The file is part of the pkg I'm trying to build: $ find . -name HsUnix.h ./include/HsUnix.h I would not expect to have to include $CWD/include normally when building a pkg but even using --extra-include-dirs does not work. I've followed the setup guide here: http://hackage.haskell.org/trac/ghc/wiki/Building/Windows/Cygwin Any help appreciated! Regards, Ben

1 0

bit of a noob question
by spot135 24 Oct '09

24 Oct '09

Ok maybe a noob question, but hopefully its an easy one. This is what I've got so far: test :: x->[a] -> (b,[b]) test x arrlist = let test1 = x a = filter (\n -> fst n == test1) arrlist test2 = map snd a in (test1, [test2]) so basically I have a list say [(a,1),(a,2),(a,3),(b,1),(b,2)] etc So I give the function a x value (a or b) in this case and it return (a,[1,2,3]) which is all gravy But, Is there a way that i dont have to supply the a or b ie i call the function and it gives me the list [(a,[1,2,3]),(b,[1,2])... I presume i need another layer of recursion but I cant figure out how to do it. Any help would be gratefully received :-) -- View this message in context: http://www.nabble.com/bit-of-a-noob-question-tp26043671p26043671.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

3 3

Re: [Haskell] ANN: quickcheck-poly - Automating QuickCheck for polymorphic/overloaded properties in a small world of types
by Henning Thielemann 24 Oct '09

24 Oct '09

Ahn, Ki Yung schrieb: > Often times QuickCheck properties are polymophic/overloaded without > giving a monomorphic type signature. But there can be several types > that you are interested. You can write scripts that check all the types > you are interested annotating the generic property for each different > monotype, which is a tedious work to do. Here, I automated this via > hint (a Haskell Interpreter Library). How it works is it first asks a > type of a property to the Haskell interpreter and gets a type string, > identifies type variables that start with lowercase alphabet, generate > all possible substitution that may or may not be a correct type, stick > the type string to the property name as a signature and collect only the > meaningful types, and then finally run quickCheck over each meaningful > types. The example below clearly illustrates what it does. Maybe you could write a block of quickCheck tests, that shall be run on different types. tests :: a -> IO () tests x = do quickCheck (\y -> law1 (y `asTypeOf x)) ... allTests = tests (undefined :: Int) tests (undefined :: Double)

1 1

A 3-line program that does not work
by Philippos Apolinarius 24 Oct '09

24 Oct '09

I have a friend who is an architect. I asked her why she does not use Haskell, since she is fond of functional programming. She writes her scripts in Clean, and needs to compile them before using them to generate postscript diagrams. In Haskell, I told her, she could use runghc, and skip the compilation step. She told me that she would consider switching to Haskell, and skipping the compilation step, if I could tell her how to write "façade" in Haskell. C:\ghc\hastex>runghc tudin.hs tudin.hs:10:19: lexical error in string/character literal (UTF-8 decoding error) After browsing the Internet, I noticed that a many of people are having the same problem. Could someone tell me what is wrong with my friend's program? import System.IO main= do outh <- openFile "garb.tsm" WriteMode hPutStrLn outh "A façade is the exterior of a building" hClose outh I would appreciate a "normal" solution, that is, I would like to type the text in any editor, or generate it with LaTeX macros, and compile it using ghc. __________________________________________________________________ Make your browsing faster, safer, and easier with the new Internet Explorer® 8. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/

4 3

Re: [Haskell-beginners] using quickcheck for blackbox testing for 3rd party apps.
by Daniel Fischer 24 Oct '09

24 Oct '09

Am Dienstag 13 Oktober 2009 18:04:52 schrieb Brent Yorgey: > Brent > > * Some smart-alecks might pipe up with something about unsafePerformIO > here. But that's not a cure, it's more like performing an emergency > tracheotomy with a ballpoint pen. Quote of the month!

2 1

STM semantics
by Paolino 24 Oct '09

24 Oct '09

I have a doubt that this code works like I want because actual STM implementation exposes its effects import Control.Concurrent import Control.Concurrent.STM main = do c <- atomically $ newTChan :: IO (TChan ()) r <- atomically $ newTVar False forkIO $ do atomically $ do r0 <- readTVar r case r0 of True -> return () False -> readTChan c myThreadId >>= killThread threadDelay 1000000 atomically (writeTVar r True) The thread stops on readTChan, but exits when I change the TVar, which happens before the readTChan. Should I trust this is the correct STM behaviour , and will not change in different implementations ? thanks paolino

3 3