March 2011 - Haskell-Cafe

Haskell Weekly News: Issue 171
by Daniel Santa Cruz 02 Mar '11

02 Mar '11

Welcome to issue 171 of the HWN, a newsletter covering developments in the [1]Haskell community. This release covers the week of February 20 - 26, 2011 . Announcements Eric Kow [2]announced that the Darcs project is staring to organize the next hacking sprint, which will be "most likely in Paris in this spring. If you fancy a weekend of hacking Haskell on a really neat real world project, come join...!" Jose Pedro Magalhaes [3]informed us of the initial release of htime, a library to mimic the unix "time" utility. Iain Alexander [4]annouced the initial release of data-type, function-combine, and flexiwrap. "data-type contains a number of fundamental types and classes useful for type-level programming, such as type-level booleans, natural numbers, and lists." I'd like to once again recognize some of the members of the community that spend their time fielding questions from the new commers. If you haven't been in the #haskell irc channel, you are just missing out! There are also those who spend their time writting great answers to the many questions in haskell's StackOverflow tag. You are what make this community an inviting place for all the people interested in Haskell. Keep it up! Quotes of the Week kmc: but rarely do beginners come to #haskell and say "i want to append two lists, but i don't understand all this Monoid business, i hear they're like taco salad" Eduard:_Munteanu [In response to "GHC can go jump out a window" and "GHC has already jumped out a window and flied and left you behind"] Yes, GHC even implements optimizations such as defenestration. Top Reddit Stories * What kind of things are easy in Haskell and hard in Scala, and vice versa?: From (programmers.stackexchange.com) scored 47 with 34 comments. Read on [5]reddit or the [6]original post. * Multiday Debugging GHC: From (blog.ezyang.com) scored 37 with 2 comments. Read on [7]reddit or the [8]original post. * Making OpenCL Simple with Haskell [pdf]: From (developer.amd.com) scored 35 with 11 comments. Read on [9]reddit or the [10]original post. * GHC 7.0.2 RC 2 available. Please test!: From (haskell.org) scored 33 with 5 comments. Read on [11]reddit or the [12]original post. * Minecraft data API for Haskell: HaskellNBT: From (blog.acfoltzer.net) scored 27 with 4 comments. Read on [13]reddit or the [14]original post. * Dependently Typed Programming: an Agda Introduction: From (youtube.com) scored 25 with 5 comments. Read on [15]reddit or the [16]original post. * Minecraft network protocol and proxy in Haskell: From (github.com) scored 25 with 1 comments. Read on [17]reddit or the [18]original post. * Threading State - a short tutorial for new Haskell programmers: From (mtnviewmark.wordpress.com) scored 22 with 3 comments. Read on [19]reddit or the [20]original post. * Solution (code) XCKD: Nerd Snipping: From (pgraycode.wordpress.com) scored 20 with 8 comments. Read on [21]reddit or the [22]original post. * Agda 2.2 released: From (permalink.gmane.org) scored 18 with 1 comments. Read on [23]reddit or the [24]original post. Top StackOverflow Answers * [25]Feed elements of a tuple to a function as arguments in Haskell? votes: 14 Function uncurry converts a two-argument (curried) function into a function on pairs. Here's its type signature: uncurry :: (a -> b -> c) -> (a, b) -> c You need to use it on printf, like this: mapM_ (uncurry $ printf "Values: %d %d\n") [(1,100),(2,350),(3,600),(4,200)] Another solution is to use pattern matching to deconstruct the tuple, ... * [26]What is the difference between different orderings of the same monad transformers? votes: 10 Edit: I originally got the cases backwards. Fixed now. The difference between orderings of monad transformer stacks really only matters when you're peeling off layers of the stack. type Procedure a = MaybeT (State ProcedureState) a In this case, you first run the MaybeT, which results in a stateful computation which returns a Maybe a. type Procedure a ... Top StackOverflow Questions * Real-world applications of zygohistomorphic prepromorphisms (votes: 54, answers: 2) [27]read * How to roll a fast BVH representation in Haskell (votes: 10, answers: 2) [28]read * curious about how âloop = loopâ is evaluated in Haskell (votes: 9, answers: 2) [29]read * What is the difference between different orderings of the same monad transformers? (votes: 8, answers: 3) [30]read * haskell function declaration (votes: 6, answers: 3) [31]read About the Haskell Weekly News To help create new editions of this newsletter, please send stories to dstcruz(a)gmail.com. I'm in dire need of finding good "quotes of the week". If you happen to come across any, please don't hesitate to send it along. Until next time, Daniel Santa Cruz References 1. http://haskell.org/ 2. http://article.gmane.org/gmane.comp.lang.haskell.cafe/86711 3. http://article.gmane.org/gmane.comp.lang.haskell.cafe/86731 4. http://article.gmane.org/gmane.comp.lang.haskell.general/18526 5. http://www.reddit.com/r/haskell/comments/ft6eh/what_kind_of_things_are_easy… 6. http://programmers.stackexchange.com/questions/51245/what-kind-of-things-ar… 7. http://www.reddit.com/r/haskell/comments/fsp4f/multiday_debugging_ghc/ 8. http://blog.ezyang.com/2011/02/multi-day-debugging/ 9. http://www.reddit.com/r/haskell/comments/fqmmu/making_opencl_simple_with_ha… 10. http://developer.amd.com/zones/OpenCLZone/publications/assets/MakingOpenCLS… 11. http://www.reddit.com/r/haskell/comments/fp5e8/ghc_702_rc_2_available_pleas… 12. http://www.haskell.org/pipermail/glasgow-haskell-users/2011-February/020056… 13. http://www.reddit.com/r/haskell/comments/fs2nw/minecraft_data_api_for_haske… 14. http://blog.acfoltzer.net/2010/12/minecraft-data-api-for-haskell-haskellnbt/ 15. http://www.reddit.com/r/haskell/comments/fr5c1/dependently_typed_programmin… 16. http://www.youtube.com/view_play_list?p=44F162A8B8CB7C87 17. http://www.reddit.com/r/haskell/comments/fsgaq/minecraft_network_protocol_a… 18. https://github.com/glguy/minecraft-proxy 19. http://www.reddit.com/r/haskell/comments/fq4k4/threading_state_a_short_tuto… 20. http://mtnviewmark.wordpress.com/2011/02/21/threading-state/ 21. http://www.reddit.com/r/haskell/comments/fr9v7/solution_code_xckd_nerd_snip… 22. http://pgraycode.wordpress.com/2011/02/23/solving-xkcds-nerd-snipping-probl… 23. http://www.reddit.com/r/haskell/comments/fp45s/agda_22_released/ 24. http://permalink.gmane.org/gmane.comp.lang.agda/2566 25. http://stackoverflow.com/questions/5117948/feed-elements-of-a-tuple-to-a-fu… 26. http://stackoverflow.com/questions/5075621/what-is-the-difference-between-d… 27. http://stackoverflow.com/questions/5057136/real-world-applications-of-zygoh… 28. http://stackoverflow.com/questions/5077243/how-to-roll-a-fast-bvh-represent… 29. http://stackoverflow.com/questions/5126759/curious-about-how-loop-loop-is-e… 30. http://stackoverflow.com/questions/5075621/what-is-the-difference-between-d… 31. http://stackoverflow.com/questions/5073058/haskell-function-declaration

1 0

Rank-2 types in classes
by Yves Parès 02 Mar '11

02 Mar '11

Hello, I'm working on a library which aims to be a generic interface for 2D rendering. To do that, one of my goals is to enable each implementation of this interface to run in its own monad (most of the time an overlay to IO), thus giving me the following class class (Monad (IM i x)) => Impl i x where data IM i x :: * -> * (where IM means Implementation Monad) Here, 'x' aims at being a phantom type that ensures a type safe context for some operations. E.g., operations that need to occur in a window will have the type: IM i Window a And will be run by the function: withinWindow :: Window -> IM i Window a -> IM i x a This makes an operation that doesn't instantiate 'x' context-independent. My problem, then, is that in the definition of class Impl the type variable 'x' is useless, since every implementation must leave uninstantiated. I would like to write something like : class (forall x. Monad (IM i x)) => Impl i where data IM i :: * -> * -> * But GHC forbids me to do so. Any suggestion would be warmly welcomed.

3 4

Re: [Haskell-cafe] type safe web forms at compile time Was: Re: Ur vs Haskell
by Alberto G. Corona 02 Mar '11

02 Mar '11

> WASH is full of wonderful ideas . packed in a not so wonderful syntax. It > is worth to evolve it. > > WASH does force form safety in a similar way to Formlets > <http://hackage.haskell.org/package/formlets>: because the form and the > form read code are generated automatically by a class instance. So there is > no need for, type checking safety > > The problem comes when the form is generated and/or maintained( edited by > some people (some forms have lot of formating in the real world) while the > form handling code is .maintained by some other people, the programmers. > > In this real case , type cheching is very important. > > 2011/3/2 Marc Weber <marco-oweber(a)gmx.de> > > Excerpts from Alberto G. Corona's message of Wed Mar 02 20:53:28 +0000 >> 2011: >> > Some time ago I forgot to forward this message to thie ur versus Haskell >> > <http://www.haskell.org/pipermail/haskell-cafe/2011-January/088060.html >> >discussion, >> > (as usual) >> > --- >> > The most impressive feature (of ur) is the compile time checking of >> > conformance between the form and the form results. This can be done in >> >> See WASH (-> hackage). So there is a Haskell implementation. There are >> small >> differences though: urweb has nicer URLS which should be much more SEO >> friendly. >> >> From my point of view its not only about forms. Its also about checking >> SQL queries. And urweb seems to do this very well. >> >> Marc Weber >> > >

1 0

type safe web forms at compile time Was: Re: Ur vs Haskell
by Alberto G. Corona 02 Mar '11

02 Mar '11

Hi. Some time ago I forgot to forward this message to thie ur versus Haskell <http://www.haskell.org/pipermail/haskell-cafe/2011-January/088060.html>discussion, (as usual) --- The most impressive feature (of ur) is the compile time checking of conformance between the form and the form results. This can be done in Haskell using HList magic and some class instances, I guess. ---- Since then I have been playing mentally with this. Recently I found something simple an interesting enough to share. (Although crude). It is a kind of typed form fields data Input a= Input String Type a (String -> Either String a) and a kind of heterogeneous list to aggregate form fields and results with the operator (:*): Input a :* Input b ;* Input c.... a :* b :* c and a (simulated for the purpose of demonstration) send-receive function that type match the form fields and the results: *Main> let form = Input "" Text "stringdata" novalidate :* Input "" Text (1::Integer) novalidate *Main> ask form >>= \(a :* b) -> return $ a ++ b <interactive>:1:0: No instance for (FormDigest (Input [Char] :* Input Integer) ([a] :* [a])) ...... notifying that there is no translation defined , because the result requires two lists of the same type when the form gives a string and an Integer But forcing the correct monomorphic types it does pattern match and return the values. *Main> ask form >>= \ (a :* b) -> print ('s':a) >> print ( fromInteger $ b) "sstringdata" 1 ask is just a simulation of HTTP one time interaction. It returns the input values. The whole loop involves the rendering of the form, with render: *Main> render form <input type="Text" name="var1" value="stringdata"/> <input type="Text" name="var2" value=1/> In a real case the results are read and validated from the the post values.They are (or can be) ordered sequentially acording with Input field names. The FormDigest instances do this work. There is no need to define new FormDigest instances. (although non one to one field-result can be created) The text is in literate haskell. There is a more elaborate example at the end. I know that the instances are non tail recursive and there are factorization pending but this is just a proof of concept: > {-# LANGUAGE > FlexibleInstances, > MultiParamTypeClasses, > TypeOperators > #-} > import Control.Monad.State The Heterogeneous list agregator. I tried to use GADTs but they does not easily pattern match > data x :* xs = x :* xs deriving (Read, Show, Eq) > infixr 5 :* > data Type= Hidden | Text deriving (Show, Read) -- to add all the types the input field, with text formatting, initial value and runtime validator > data Input a = Input String Type a (String -> Either [String] a) > instance(Show a)=> Show (Input a) where > show (Input _ _ x _) = show x rendering of the form need a sequentiation of field names. I use a state monad for this > class RenderForm a where > renderForm :: a -> State Int String > instance (Show a) => RenderForm (Input a) where > renderForm input = do > s1 <- render1 input > n <- get > put $ n + 1 > return s1 HList school here: > instance (Show a,RenderForm xs) => RenderForm (Input a :* xs) where > renderForm (input :* xs)= do > n <- get > put $ n+1 > h <- render1 input > s <- renderForm xs > return $ h++s > render1 (Input msg t x _)= do > n <- get > put $ n+1 > return $ msg > ++ "<input type=\"" > ++ show t ++ "\" name=" > ++ "\"var"++show n ++ "\" value=" > ++ show x ++"/>\n" > > render form= putStrLn $ evalState (renderForm form ) 0 processing of the returned form result, in an ordered String list, according with seuquential names of the fields defined in renderForm. > class FormDigest a b where > formDigest :: a -> [String] -> Either [String] b "Input a" is diggested into a type "a" > instance FormDigest (Input a) (a) where > formDigest (Input _ _ x f) (s: ss)= case f s of > Right x -> Right $ x > Left x -> Left x recursively add instances for any list of inputs Input a's are diggested into a's > instance FormDigest as bs > => FormDigest (Input a :* as) (a :* bs) where > formDigest (input :* fs) es@(s:ss) = > let er = formDigest fs ss > e = formDigest input es > in case (e, er) of > (Right x, Right ys) -> Right $ x :* ys > (Right _, Left errs) -> Left errs > (Left err, Left errs) -> Left (err ++ errs) simulated request-response that returns the entered input values > sendRec xs= do > let strs = showValues xs > return $ formDigest xs strs > class ShowValues a where > showValues :: a -> [String] > instance Show x => ShowValues (Input x) where > showValues i = [show i ] > instance (Show x,ShowValues xs) => ShowValues (Input x :* xs) where > showValues (i :* xs)= show i : showValues xs end of simulated request response > ask :: (ShowValues a, FormDigest a b) => a -> IO b > ask form = do > er <- sendRec form > case er of > Left errs -> error "" -- shoud be: "ask1 errs form "·to render form and errors > Right res -> return res EXAMPLE: > data Emp= Emp{name::String, salary :: Int} deriving Show > emp= Emp "Luis" 10000 toy html operators: > b msg = ("<b> " ++ msg ++ " </b>\n") > p msg = ("<p> " ++ msg ++ " </p>\n") > novalidate n= Right $ read n > main= do > > let form = Input > ( b "please enter the name" > ++ p "mas texto") > Text (name emp) novalidate > :* Input > (b "please enter the salary" > ++ p "jkjkjk") > Text (salary emp) novalidate > render form the matching thing > (n :* s ) <- ask form > print emp > print $ Emp n s

2 1

Examples for the problem
by Robert Clausecker 02 Mar '11

02 Mar '11

Thank you all for the responses. Here's an example: As I alrerady said, I tried to parse the MMIXAL assembly language. Each instruction has up to three operands, looking like this: @+4 (Jump for bytes forward) "foo" (the string foo" '0'>>(1+2) etc. A string literal may contain anything but a newline, (there are no escape codes or similar). But when I add a check for a newline, the parser just fails and the next one is tried. This is undesired, as I want to return an error like "unexpected newline" instead. How is this handled in other parsers?

5 5

HaXml combinators
by DCE 01 Mar '11

01 Mar '11

Hello, I'm trying to use HaXml to do some relatively simple XML processing, but I can't figure out how to get started. My documents look something like this: <doc> <a bar="x"> <b foo="x"/> <b foo="y"/> <a bar="y"> <b foo="z"/> </a> </a> </doc> I want to find any element tagged <a> that has an attribute bar=y, then find the element under it tagged <b> with attribute foo=z, then add an attribute "baz=t" to this <b>. I've tried various incantations like (tag "b" `o` children `o` (tag "a" `with` attrval ...)), but they don't seem to work. And in general, I'm not sure how to accomplish looking up and down the XML tree from a given node. I have figured out how to use mkElemAttr to build a new element, just not how to find the correct elements in the first place. Additionally, I'm confused about when to use foldXml, deep, multi, and so on for processing elements, and whether the 'inplace' combinator is needed for replacing the <b> element with the result of mkElemAttr. Any help would be appreciated.

1 0

SMP parallelism gains inferior than expected
by José Pedro Magalhães 01 Mar '11

01 Mar '11

(Forwarding to haskell-cafe) Hi, I have a program that computes a matrix of Floats of m rows by n columns. Computing each Float is relatively expensive. Each line is completely independent of the others, so I thought I'd try some simple SMP parallelism on this code: myFun :: FilePath -> IO () myFun fp = do fs <- readDataDir fp let process f = readFile' f >>= parse printLine = putStrLn . foldr (\a b -> show a ++ "\t" ++ b) "" runDiff l = [ [ diff x y | y <- l ] | (x,i) <- zip l (map getId fs), myFilter i ] ps <- mapM process fs sequence_ [ printLine x | x <- runDiff ps *`using` parList rdeepseq* ] So, I'm using parList to evaluate the rows in parallel, and fully evaluating each row. Here are the timings on a Dual Quad Core AMD 2378 @2.4 GHz, ghc-6.12.3, parallel-2.2.0.1: -N time (ms) none 1m50 2 1m33 3 1m35 4 1m22 5 1m11 6 1m06 7 1m45 The increase at 7 is justified by the fact that there were two other processes running. I don't know how to justify the small increase at N3, though, but that doesn't matter too much. The problem is that I am not getting the gains I expected (halving at N2, a third at N3, etc.). Is this the best one can achieve with this implicit parallelism, or am I doing something wrong? In particular, is the way I'm printing the results at the end destroying potential parallel gains? Any insights on this are appreciated. Thanks, Pedro

2 1

Rebindable 'let' for observable sharing?
by Henning Thielemann 01 Mar '11

01 Mar '11

Now that almost every syntax can be redirected to custom functions (RebindableSyntax, OverloadedStrings), would it make sense to map 'let' to 'fix' ? Would this open a way to implement observable sharing as needed in EDSLs by a custom 'fix'?

3 2

segfault when using ghc api
by Edward Amsden 28 Feb '11

28 Feb '11

I'm trying to run the following code. I'm not at all sure it's correct, it's based off of a bit of poking around in the ghc api. Running it with a command line argument like "show (5 + 2)" gives me a segmentation fault. Poking around with gdb and following the steps at http://hackage.haskell.org/trac/ghc/wiki/Debugging/CompiledCode yields precisely nothing, as even the "disassemble" command complains "No function contains program counter for selected frame." Any ideas? == code == module Main where import GHC import DynFlags import Data.Dynamic import System evalString :: Typeable a => String -> IO (Maybe a) evalString s = defaultErrorHandler defaultDynFlags $ runGhc Nothing $ do dynflags <- getSessionDynFlags setSessionDynFlags $ dynflags target <- guessTarget "Prelude" Nothing setTargets [target] load LoadAllTargets dyn <- dynCompileExpr s return $ fromDynamic dyn main = do (s:_) <- getArgs e <- evalString s putStrLn $ maybe "oops" id e -- Edward Amsden Undergraduate Computer Science Rochester Institute of Technology www.edwardamsden.com

3 4

HANSEI in Haskell?
by Daryoush Mehrtash 28 Feb '11

28 Feb '11

Is the "Embedded domain-specific language HANSEI for probabilistic models and (nested) inference" described in: http://okmij.org/ftp/kakuritu/index.html#implementation available in Haskell? Is there a reason why the author did the package in Ocaml rather than Haskell? -- Daryoush

3 8