- Haskell-Cafe - Haskell.org

Stacking data types
by Yves Parès 11 Apr '11

11 Apr '11

Hello Café, I'm trying to get some modular data types. The idea that came to me is that I could stack them, for instance : data Character a = Character { life :: Int, charaInner :: a } data Gun a = Gun { firepower :: Int, gunInner :: a } data Armor a = Armor { resistance :: Int, armorInner :: a } Then a character with a gun and an armor can be build this way: chara = Character 100 $ Armor 40 $ Gun 12 The idea now is to be able to get some part of the character: itsGun :: Character ?? -> Gun ?? itsGun = content Then content would be a class method: class Has b a where content :: a -> b And it would be recursively defined so that: instance (Has c b, Has b a) => Has c a where content = (content :: b -> c) . (content :: a -> b) Then itsGun would be more like: itsGun :: (Has Gun a) => a -> Gun ?? itsGun = content But after some juggling with extensions (ScopedTypeVariables, UndecidableInstances, IncoherentInstances...) I can't get it working. Has someone a simpler way to achieve modular types?

6 8

Foreign function performance: monadic vs pure
by Serguei Son 11 Apr '11

11 Apr '11

Consider two versions of sin wrapped: foreign import ccall "math.h sin" c_sin_m :: CDouble -> IO CDouble and foreign import ccall "math.h sin" c_sin :: CDouble -> CDouble One can invoke them so: mapM c_sin_m [1..n] mapM (return . c_sin) [1..n] On my computer with n = 10^7 the first version never finishes, whereas the second one calculates the result within seconds. To give you my context, I need to call a random variable generator multiple times, so that it must return IO a. Any explanation for this behavior?

5 7

Emscripten: compiling LLVM to JavaScript
by Sönke Hahn 11 Apr '11

11 Apr '11

Hi all! I haven't tried the tool myself, but it seems interesting to the Haskell efforts to compile to JavaScript: http://syntensity.blogspot.com/2011/04/emscripten-10.html Cheers, Sönke

6 5

Re: [Haskell-cafe] already installed packages alerted as not being installed
by Kazu Yamamoto 11 Apr '11

11 Apr '11

Hello, > thanks I wanted to mention that the "unknown symbol" error is very > likely not related to the cab tool as the same error appears, when > using the cabal - tool. I guess we can ignore it even in the context > of my main question, sorry for being to verbose. What I found more > interesting is, that cab doesn't try to re-install an installed > package - already in the database too, as ghc-pkg output pretends - > which is expected behaviour. The cabal system frontend obviously does > trying to reinstall the package though. This seems like different > model of reality... This is not true. cab install re-installs some packages if necessary, unfortunately. That's why I recommend to use "cab install -n". --Kazu

2 1

how to optmize this code?
by Gilberto Garcia 11 Apr '11

11 Apr '11

Hi Haskellers, I was solving this problem from project euler to study haskell. I came up whit the following solution and I was wondering if there is a more optimized and concise solution. fkSum :: Int -> [Int] -> Int fkSum a [] = 0 fkSum a (b) = foldl (+) 0 (filter (\x -> isMultiple x b) [1..a]) isMultiple :: Int -> [Int] -> Bool isMultiple a [] = False isMultiple a (x:xs) = if (mod a x == 0) then True else isMultiple a xs Thanks in advance ggarcia

8 12

Re: [Haskell-cafe] Fucntion composing
by Adam Krauze 11 Apr '11

11 Apr '11

Thanks all! You helped me a lot. Adam.

1 0

Parsing HTML tables with HXT
by Dmitry Simonchik 11 Apr '11

11 Apr '11

Dear Cafe, Can someone please help me with getting the value of the table cell with HXT in the following html: <table class="tblc"> <tr> <td class="tdc">x</td> <td>y</td> </tr> <tr> <td class="tdc">a</td> <td>b</td> </tr> </table> I need the value of the second cell in a row that has first cell with some predefined value (in the example above it can be x or a) I need the arrow of the type (IOSArrow XmlTree String) How to write it? Thanks

2 2

already installed packages alerted as not being installed
by Daniel Kahlenberg 11 Apr '11

11 Apr '11

Hello cafe-readers, does anyone of you observe similar problems e. g. on a Windows with ghc-7.0.2 setup: When I'm trying cabal install threadscope (as an example package depending on gtk2hs, latter which I've installed using the stepwise approach of cabal unpack first, then cabal configure --user etc.) it complains about cairo etc. not being installed and tries to re-install it on-the-fly, failing due to whats described her: http://hackage.haskell.org/trac/gtk2hs/ticket/1203! When I install cabal-dev and cab first and then re-install everything with cab instead of cabal the issue with re-installing already installed packages described above disappears and only an "unknown symbol" message related to the correctly found installed cairo package remains. So is there an error in package database handling somewhere or changed semantics in cabal | ghc-pkg | (even) pkg-config flags I missed? Cheers Daniel

2 2

CFP: APLAS 2011, Kenting, Taiwan
by Shin-Cheng Mu 11 Apr '11

11 Apr '11

========================================================== APLAS 2011 Call For Papers Ninth Asian Symposium on Programming Languages and Systems Kenting, Taiwan, December 5--7, 2011 (co-located with CPP 2011) http://flolac.iis.sinica.edu.tw/aplas11/ ========================================================== BACKGROUND: APLAS aims at stimulating programming language research by providing a forum for the presentation of latest results and the exchange of ideas in topics concerned with programming languages and systems. APLAS is based in Asia, but is an international forum that serves the worldwide programming language community. APLAS is sponsored by the Asian Association for Foundation of Software (AAFS) founded by Asian researchers in cooperation with many researchers from Europe and the USA. The past APLAS symposiums were successfully held in Shanghai ('10), Seoul ('09), Bangalore ('08), Singapore ('07), Sydney ('06), Tsukuba ('05), Taipei ('04) and Beijing ('03) after three informal workshops. Proceedings of the past symposiums were published in Springer-Verlag's LNCS 6461, 5904, 5356, 4807, 4279, 3780, 3302, and 2895. The 2011 edition will be held at Kenting, a seaside resort in Southern Taiwan. TOPICS: The symposium is devoted to both foundational and practical issues in programming languages and systems. Papers are solicited on, but not limited to, the following topics: + semantics, logics, foundational theory; + design of languages and foundational calculi; + type systems; + compilers, interpreters, abstract machines; + program derivation and transformation; + program analysis, verification, model-checking, software security; + concurrency, constraints, domain-specific languages; + tools for programming, verification, implementation. APLAS 2011 is not limited to topics discussed in previous symposiums. Papers identifying future directions of programming and those addressing the rapid changes of the underlying computing platforms are especially welcome. Demonstration of systems and tools in the scope of APLAS are welcome to the System and Tool presentations category. Authors concerned about the appropriateness of a topic are welcome to consult with Program Chair prior to submission. IMPORTANT DATES: Abstract Deadline: June 13, 2011 Submission Deadline: June 17, 2011 (GMT) Notification: August 29, 2011 Camera-Ready: September 19, 2011 INVITED SPEAKERS: Nikolaj Bjorner, Microsoft Research Redmond Ranjit Jhala, UC San Diego Peter O'Hearn, Queen Mary U. London Sriram Rajamani, Microsoft Research India SUBMISSION INFORMATION: We solicit submissions in two categories: Regular research papers, describing original research results, including tool development and case studies, from a perspective of scientific research. Regular research papers should not exceed 16 pages in the Springer LNCS format, including bibliography and figures. They should clearly identify what has been accomplished and why it is significant. Submissions will be judged on the basis of significance, relevance, correctness, originality, and clarity. In case of lack of space, proofs, experimental results, or any information supporting the technical results of the paper could be provided as Appendix or a link to a web page. System and Tool presentations, describing systems or tools that support theory, program construction, reasoning, and/or program execution in the scope of APLAS. Unlike presentations of regular research papers, presentation of accepted papers in this category is expected to be centered around a demonstration. The paper and the demonstration should identify the novelties of the tools and use motivating examples. System and Tool papers should not exceed 8 pages in the Springer LNCS format, including bibliography and figures. Submissions will be judged based on both the papers and the systems or tools as described in the papers. It is highly desirable that the tools are available on the web. Papers should be submitted electronically via the submission web page at http://www.easychair.org/conferences/?conf=aplas2011. Acceptable formats are PostScript or PDF, viewable by Ghostview or Acrobat Reader. Submitted papers must be unpublished and not submitted for publication elsewhere. Papers must be written in English. The proceedings are planned to be published as a volume in Springer-Verlag's Lecture Notes in Computer Science series. Accepted papers must be presented at the conference. GENERAL CHAIR: Tyng-Ruey Chuang, Academia Sinica, Taiwan PROGRAM CHAIR: Hongseok Yang, Queen Mary U. London aplas2011(a)easychair.org PROGRAM COMMITTEE: Lars Birkedal (ITU, Denmark) James Brotherston (Imperial College, UK) Kung Chen (National Chengchi U., Taiwan) Wenguang Chen (Tsinghua U., China) Wei-Ngan Chin (NUS, Singapore) Javier Esparza (TUM, Germany) Xinyu Feng (USTC, China) Jerome Feret (INRIA, France) Matthew Fluet (RIT, USA) Rajiv Gupta (UC Riverside, USA) Masahito Hasegawa (Kyoto U., Japan) Radha Jagadeesan (Depaul U., USA) Naoki Kobayashi (Tohoku U., Japan) Daniel Kroning (U. Oxford, UK) Rupak Majumdar (MPI-SWS, Germany) Andrzej Murawski (U. Leicester, UK) Paulo Oliva (Queen Mary U. London, UK) Doron Peled (Bar Ilan U., Israel) Sukyoung Ryu (KAIST, Korea) Sriram Sankaranarayanan (U. Colorado, USA) Armando Solar-Lezama (MIT, USA) Sam Staton (U. Cambridge, UK) Viktor Vafeiadis (MPI-SWS, Germany) Kapil Vaswani (MSR, India) Martin Vechev (IBM, USA) Peng Wu (IBM, USA) Hongseok Yang (Queen Mary U. London, UK) Pen-Chung Yew (Academia Sinica, Taiwan) ORGANIZERS: Mike Dodds, U. Cambridge (Poster chair) Shin-Cheng Mu, Academia Sinica, Taiwan (Local arrangement chair) Noam Rinetzky, Queen Mary U. London (Publicity chair) Yih-Kuen Tsay, National Taiwan University (Finance chair)

1 0

Re: [Haskell-cafe] [Haskell-beginners] Calling a foreign function: superlinear comlexity
by Antoine Latter 11 Apr '11

11 Apr '11

On Sun, Apr 10, 2011 at 6:26 PM, Serguei Son <serguei.son(a)gmail.com> wrote: > I call GSL's gsl_ran_ugaussian function in the following way (using > bindings-gsl): > > module Main where > > import Bindings.Gsl.RandomNumberGeneration > import Bindings.Gsl.RandomNumberDistributions > import Foreign > import Control.Monad > import Data.List > > main = do > let n = 100000 > p <- peek p'gsl_rng_mt19937 > rng <- c'gsl_rng_alloc p > lst <- replicateM n $ c'gsl_rng_uniform rng > print $ sum lst > > As I increase n from 10^4 to 10^5 to 10^6 execution time grows superlinearly. > Not sure if it is related, but this thread also documents issues folks had with replicateM and performance: http://www.haskell.org/pipermail/haskell-cafe/2011-March/090419.html Antoine > To forestall the answer that the reason is the overhead of List, > this code scales approximately linearly: > > module Main where > > > import Foreign > import Control.Monad > import Data.List > > main = do > let n = 100000 > let lst = map sin [1..n] > print $ sum lst > > Another interesting observation: when I wrap the sin function > of math.h with signature CDouble -> IO CDouble calling it > repeatedly scales superlinearly, whereas when I wrap it as a pure > function calling it repeatedly scales linearly. > > What is the reason for this performance and how can > I make the first code scale linearly in execution time? > > > > > > _______________________________________________ > Beginners mailing list > Beginners(a)haskell.org > http://www.haskell.org/mailman/listinfo/beginners >

1 0