- Haskell-Cafe - Haskell.org

Explicit garbage collection
by Miguel Mitrofanov 07 Jan '10

07 Jan '10

Is there any kind of "ST" monad that allows to know if some STRef is no longer needed? The problem is, I want to send some data to an external storage over a network and get it back later, but I don't want to send unnecessary data. I've managed to do something like that with weak pointers, System.Mem.performGC and unsafePerformIO, but it seems to me that invoking GC every time is an overkill. Oh, and I'm ready to trade the purity of runST for that, if necessary. Thanks.

3 12

Example of sending headers/cookies with Network.HTTP.simpleHTTP or Network.Browser?
by Matt Brown 07 Jan '10

07 Jan '10

Hi all, Does anyone know of a good/simple example of sending cookies and other headers in an HTTP request via simpleHTTP or Browser? I know what the cookies should be (i.e. I don't want them to be sent by the server). More specifically, I'm looking for an analogue to curl's -b and -H flags. thanks, -matt

1 0

Distinct types in a list
by rodrigo.bonifacio 07 Jan '10

07 Jan '10

Hi all, I have a family of parsers that return either (Success t) or (Fail), using the following data type: > data ParserResult a = Success a | Fail String > deriving (Read, Show, Eq, Ord) > > isSuccess (Success _) = True > isSuccess (Fail _) = False > ... I want to add the results of different parsers to a list. Such as: > m1 = parseFirstModel file1 -- it returns a ParserResult of t1 > m2 = parseSecondModel file2 -- it returns a ParserResult of t2 > ps = [m1, m2] In such a way that I could write something like: > if and (map isSuccess ps) > then process m1 m2 > else ... Actually, in the real program I have to check more than two input models. However, since Lists do only hold elements of a same type, I couldn't proceed in this way. Which improvements to the ParserResult data type should I wrote in order to proceed as I want to. Best regards, Rodrigo.

6 5

pcre-light install fails with undefined reference to _impure_ptr
by Patrick Caldon 07 Jan '10

07 Jan '10

Not seen this one before. I'm trying to install pcre-light with WinXP 64, Cygwin 1.7.1, Cabal 0.8.0, and ghc 6.12.1. Trying to reinstall "base" with cabal fails as well with an internal error (see below), but I'm assuming that's quasi-intentional/unsupported feature or something. Any ideas? My ultimate goal is to install ghc-core to help track down some performance problems I'm having. Thanks, Patrick. patc@rose ~/rp4/loansim $ cabal install pcre-light --extra-lib-dirs=C:\\cygwin\\lib --extra-include-dirs=C:\\cygwin\\usr\\include --reinstall Resolving dependencies... Configuring pcre-light-0.3.1... Preprocessing library pcre-light-0.3.1... In file included from C:/cygwin/usr/include/pcre.h:90, from Text\Regex\PCRE\Light\Base.hsc:103: C:/cygwin/usr/include/stdlib.h:104: warning: `warning' attribute directive ignored C:/cygwin/usr/include/stdlib.h:109: warning: `warning' attribute directive ignored dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o:Base_hsc_make.c:(.text+0x47): undefined reference to `_impure_ptr' dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o:Base_hsc_make.c:(.text+0x7b): undefined reference to `_impure_ptr' dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o:Base_hsc_make.c:(.text+0x93): undefined reference to `_impure_ptr' dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o:Base_hsc_make.c:(.text+0xc7): undefined reference to `_impure_ptr' dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o:Base_hsc_make.c:(.text+0xf3): undefined reference to `_impure_ptr' dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o:Base_hsc_make.c:(.text+0x127): more undefined references to `_impure_ptr' follow collect2: ld returned 1 exit status linking dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o failed command was: C:\ghc\ghc-6.12.1\mingw\bin\gcc.exe -LC:\cygwin\lib -lpcre -LC:\ghc\ghc-6.12.1\bytestring-0.9.1.5 -LC:\ghc\ghc-6.12.1\base-3.0.3.2 -LC:\ghc\ghc-6.12.1\syb-0.1.0.2 -LC:\ghc\ghc-6.12.1\base-4.2.0.0 -lwsock32 -luser32 -lshell32 -LC:\ghc\ghc-6.12.1\integer-gmp-0.2.0.0 -LC:\ghc\ghc-6.12.1\ghc-prim-0.2.0.0 -LC:\ghc\ghc-6.12.1 -LC:\ghc\ghc-6.12.1/gcc-lib -lm -lwsock32 -LC:\ghc\ghc-6.12.1 dist\build\Text\Regex\PCRE\Light\Base_hsc_make.o -o dist\build\Text\Regex\PCRE\Light\Base_hsc_make.exe cabal.exe: Error: some packages failed to install: pcre-light-0.3.1 failed during the building phase. The exception was: exit: ExitFailure 1 patc@rose ~/rp4/loansim $ cabal --version cabal-install version 0.8.0 using version 1.8.0.2 of the Cabal library patc@rose ~/rp4/loansim $ ghc --version The Glorious Glasgow Haskell Compilation System, version 6.12.1 patc@rose ~/rp4/loansim $ cabal install base --reinstall -p Resolving dependencies... cabal.exe: internal error: impossible patc@rose ~/rp4/loansim $

3 5

Re: Why doesn't laziness save the day here?
by oleg＠okmij.org 07 Jan '10

07 Jan '10

Dale Jordan wrote: > The motivation for iterateR is to be able to have the ultimate > consumer determine how many random values to force, but still have a > single random generator used throughout the computation. > > My intuition tells me that since the infinite list is produced in > finite batches, ... > > Can anyone explain why this is looping or point out a better way to > generate an arbitrary-length random list while still being able to > reuse the generator? The others have already pointed out the problem with the imperative solution, which used the mutation of the global state with the new random seed. Imperative approach is indeed often a problem. There is a simple solution however. In fact, your message already described it. The key phrase is ``the infinite list is produced in finite batches.'' We merely need to define a list-like data structure that matches our intuitions. The complete code follows. All three run functions, run1 through run3, produce a finite result (shown in the comments after the definition). module RList where -- I don't have Mersenne Twistor installed; so I'll use the stdGen... import System.Random import Control.Monad.State data RList m a = RList [a] -- known finite prefix [m [a]] -- a stream of producing actions pullR :: Monad m => RList m a -> m (RList m a) pullR (RList p (x:xs)) = x >>= \p' -> return $ RList (p++p') xs headR :: Monad m => RList m a -> m a headR (RList (x:_) _) = return x headR x = pullR x >>= headR tailR :: Monad m => RList m a -> m (RList m a) tailR (RList (_:xs) ms) = return $ RList xs ms tailR x = pullR x >>= tailR -- appendR doesn't have to have the monadic type. We go for uniformity with -- the headR appendR :: Monad m => RList m a -> RList m a -> m (RList m a) appendR (RList p1 ms1) (RList p2 ms2) = return $ RList p1 (ms1 ++ (return p2):ms2) takeR :: Monad m => Int -> RList m a -> m (RList m a) takeR 0 l = return l takeR n (RList p ms) | length p >= n = return $ RList (take n p) [] takeR n l = pullR l >>= takeR n -- quite inefficient, but short -- Other list library functions can be implemented in terms of head, tail -- the evaluator, so to speak. It is possibly strict, use it at the -- very end toList :: (Functor m, Monad m) => RList m a -> m [a] toList (RList p []) = return p toList (RList p ms) = pullR (RList [] ms) >>= fmap (p ++) . toList -- Dale Jordan's library, slightly re-written -- Specialized iterator for running actions in Rand monad, to thread -- the generator. The idea is that the action returns a finite list -- of random values and iterateR lazily creates an infinite list of -- values. -- Again, the monadic type is unncessary; given for the sake of -- uniformity iterateR :: Monad m => m [a] -> m (RList m a) iterateR act = return $ RList [] (repeat act) type Rand r a = State r a -- A simple example of a finite action something :: (RandomGen g) => Int -> Int -> Rand g [Int] something n m = sequence . replicate n . State $ randomR (m,m+9) run1 = evalState (toList =<< takeR 10 =<< (iterateR (something 2 0))) $ mkStdGen 42 -- [1,1,7,4,6,1,8,1,8,5] run2 = evalState (toList =<< takeR 10 =<< (iterateR (something 2 0) >> iterateR (something 3 10))) $ mkStdGen 42 -- [11,11,17,14,16,11,18,11,18,15] run3 = evalState (toList =<< (takeR 10 =<< (iterateR (something 2 0) >>= headR >>= iterateR . something 3))) $ mkStdGen 42 -- [8,5,7,2,9,2,9,6,6,10]

2 1

dph and ghc-6.12 ?
by Johannes Waldmann 07 Jan '10

07 Jan '10

Dear all, I'd love to use Data Parallel Haskell (for now, just as a demonstration, in a lecture). I want to convince my students (and myself) that this thing "really works". What would be an easy-to-understand, but still impressive enough benchmark? (to be run on 8 core amd64)? What is the current status of dph with ghc-6.12? Is there some up-to-date tutorial? Do we still need separate modules/compilation/linking? (... and where can I download the CUDA backend :-) best regards, J.W.

1 0

ANNOUNCE: safer-file-handles-0.1
by Bas van Dijk 07 Jan '10

07 Jan '10

Hello, I'm happy to announce another member in the 'monadic regions' family: safer-file-handles: http://hackage.haskell.org/package/safer-file-handles-0.1 The package uses my 'regions' and 'explicit-iomodes' packages to add two safety features on top of the regular System.IO file handles and operations: * Regional file handles. Files must be opened in a region. When the region terminates all opened files are automatically closed. The main advantage of regions is that the handles to the opened files can not be returned from the region which ensures no I/O with closed files is possible. * Explicit IOModes. The regional file handles are parameterized by the IOMode in which they were opened. All operations on handles explicitly specify the needed IOMode. This way it is impossible to read from a write-only handle or write to a read-only handle for example. The primary technique used in this package is called "Lightweight monadic regions" which was invented by Oleg Kiselyov and Chung-chieh Shan. See: http://okmij.org/ftp/Haskell/regions.html#light-weight This technique is implemented in the 'regions' package which is re-exported from 'safer-file-handles'. See the 'safer-file-handles-examples' package for examples how to use this package: darcs get http://code.haskell.org/~basvandijk/code/safer-file-handles-examples regards, Bas P.S. While I wrote this message I realized a serious lack of expressive power of my regions package. Namely that I can't write a program that opens two different types of resources in the _same_ region. If for example I want to combine my usb-safe package and this package the following won't type-check: openBoth ∷ Device → FilePath → IO () openBoth usbDevice filePath = runTopRegion $ do h1 ← open usbDevice ∷ RegionT Device s IO (RegionalHandle Device (RegionT Device s IO)) h2 ← openFile filePath ReadMode ∷ RegionT File s IO (RegionalHandle File (RegionT File s IO)) return () The reason, as can be seen from the types of the monadic actions, is that the monad types differ between the two actions and bind ((>>=) :: Monad m => m a -> (a -> m b) -> m b) obviously requires them to be the same. The main problem thus, is that I parameterized the region with the type of resources. Note that I'm working on a solution where I remove the resource type from RegionT. So instead of: newtype RegionT resource s (pr ∷ * → *) α = RegionT { unRegionT ∷ ReaderT (IORef [Opened resource]) pr α } I will have: newtype RegionT s (pr ∷ * → *) α = RegionT { unRegionT ∷ ReaderT (IORef [SomeOpenedResource]) pr α } data SomeOpenedResource = ∀ resource. Resource resource ⇒ Some (Opened resource) Hopefully I will be able to release this fix before FP-NL tomorrow.

1 0

Problem with cabal install zlib
by Ozgur Akgun 07 Jan '10

07 Jan '10

Hi, When I run cabal install zlib or cabal upgrade zlib I get the following error: Resolving dependencies... Configuring zlib-0.5.2.0... Preprocessing library zlib-0.5.2.0... Only one output file may be specified cabal: Error: some packages failed to install: zlib-0.5.2.0 failed during the building phase. The exception was: exit: ExitFailure 1 I have zlib-0.5.0.0 installed now. The OS is Snow Leopard. I couldn't find anything related to this error, that's why I'm asking community's opinion. Best, -- Ozgur Akgun

6 15

ANNOUNCE: tuntap-0.0.1
by John Van Enk 07 Jan '10

07 Jan '10

Hi All, As some of you have requested, I've extracted the code from my (under work) VPN which talks to the TUN/TAP device under Linux (/dev/net/tun). Some things to note: 1) I've only tested this under Linux. 2) The help-win.c file under cbits HAS allowed similar code to talk to the TUN/TAP driver used by the OpenVPN project, but I haven't implemented this functionality yet. 3) The help-bsd.c file under cbits HAS allowed Mac/BSD flavors to talk to a TUN/TAP driver, but I've never tested this and haven't implemented this functionality yet. I don't have a whole lot there right now, but keep a lookout for more releases with some examples and better docs (as I get around to it). For a very verbose example, you can look at how I use the Network.TUNTAP module in this project: http://github.com/sw17ch/Scurry, specifically http://github.com/sw17ch/Scurry/blob/master/src/Scurry/TAPTask.hs. Hackage: http://hackage.haskell.org/package/tuntap GitHub: http://github.com/sw17ch/tuntap John Van Enk PS: Thanks to Matthew Isleb and Job Vranish for their help on the C code used.

1 0

Value classes
by Mads Lindstrøm 07 Jan '10

07 Jan '10

Hi A function inc5 inc5 :: Float -> Float inc5 x = x + 5 can only be a member of a type class A, if we make all functions from Float -> Float members of type class A. Thus, I assume, that _type_ class is named type class, as membership is decided by types. However, it makes no sense to say that all functions from Float -> Float are invertible or continuous. We would want to specifically say that inc5 is continuous, not all Float -> Float functions. Thus, we could have another abstraction, lets call it _value_ class, where we could say that an individual value (function) could be member. We could say something like: class Invertible (f :: a -> a) where invert :: f -> (a -> a) instance Invertible inc5 where invert _ = \x -> x - 5 In many cases this would be too specific. We would like to say, that applying the first argument to `+` returns an invertible function. Something like: instance Invertible (`+` x) where invert (x +) = \y -> y - x We would properly also like to say, that composing two invertible functions results in another invertible function. I guess there are many more examples. This idea, of value classes, do not feel all that novel. Somebody has properly thought about it before, but gave it a different name. If anybody has links to some papers it would be much appreciated. If anybody has some thoughts of the desirability of value class it would also be much appreciated. /Mads Lindstrøm

3 2