November 2008 - Haskell-Cafe

Hackage web interface
by Hugo Pacheco 12 Nov '08

12 Nov '08

Hi, When previewing some package via the Hackage web interface, I get the following warning: Exposed modules use unallocated top-level names: AI assume that if you define some module A.B, it expects the top-level name A to be a module. Does this make sense? Thanks, hugo -- www.di.uminho.pt/~hpacheco

2 1

problem with boolean splicing in templates in ghc 6.10.1
by Leonidas Fegaras 12 Nov '08

12 Nov '08

Seems that boolean splicing in haskell templates in ghc 6.10.1 does not work correctly. If you do: $((\b -> [| b |]) True) you get the error: Can't find interface-file declaration for data constructor GHC.Base.True Probable cause: bug in .hi-boot file, or inconsistent .hi file Use -ddump-if-trace to get an idea of which file caused the error It seems that GHC.Base.True is now GHC.Bool.True, but the splicing code still generates the former. Leonidas Fegaras

2 1

reliable (bi)directional pipe to a process
by oleg＠okmij.org 12 Nov '08

12 Nov '08

I'd like to point out a reliable, proven and simple way of interacting with another process, via unidirectional or bidirectional pipes. The method supports Unix sockets, pipes, and TCP sockets. I too have noticed insidious bugs in GHC run-time when communicating with another process via a pipe. I tried to use runInteractiveProcess; it worked -- up to file sizes of about 300Kb. Then GHC run-time seems to `loses synchronization' -- and corrupts IO buffers, receiving stuff that cannot have been possibly sent. This is because handle operations are asynchronous and the GHC scheduler seems to have race conditions. That behavior was totally unacceptable. I was writing a production code, and can't afford such errors. Therefore, I wrote a simple foreign function interface to the code sys_open that I have been using for about fifteen years. This code does work, in production, for very large file sizes and long-running processes, on many Unix and Unix-like systems. I was told once about a Cygwin port. http://okmij.org/ftp/syscall-interpose.html#Application http://okmij.org/ftp/packages/sys_open.c http://okmij.org/ftp/Haskell/MySysOpen.hs Please see the test at the end of the file MySysOpen.hs. The test interacts with another process over a bi-directional pipe, repeatedly sending and receiving data. The amount of received data is large (about 510K).

5 6

ByteString/parsec
by Pieter Laeremans 12 Nov '08

12 Nov '08

HI, I want to efficiently parse a large collections of files. The files are in the format : example title TITLE author name AUTHOR some lines with summary here SUMMARY the real text TEXT a list of links LINKS I want to use "ByteString" here, but which library should I use to parse ? "attoparsec" or "bytestringparser", both export the same interface. When I use one of these I thaught it would be nice to write something like this : fileParser :: Parser Content fileParser = do title <- manyTill getInput (string . pack "\nTITLE\n") author <- manyTill getInput (string. pack "\nTITLE\n") .... return Content title author ... But this doesn't work. Even on a small example : parseTest (manyTill getInput (string $ pack "SPLIT") (pack "split the text at SPLIT part two") I get a stack overflow. Obviously I'm not understanding something here. Are there any good examples of open source projects which parse ByteString data ? thanks in advance, Pieter -- Pieter Laeremans <pieter(a)laeremans.org> "The future is here. It's just not evenly distributed yet." W. Gibson

3 2

compiling haskell-src-exts in 6.10
by Hugo Pacheco 12 Nov '08

12 Nov '08

When installing package haskell-src-exts via cabal install, I get the error Language/Haskell/Exts/Syntax.hs:102:7: Could not find module `Data.Data': it is a member of package base, which is hidden However, when manually installing runhaskell Setup.hs configure/build/install It works fine Somehow package base is not available in cabal.. don't know why. hugo -- www.di.uminho.pt/~hpacheco

3 2

Re[2]: [Haskell-cafe] mapping unfreeze over an IntMap of IOUArrays
by Bulat Ziganshin 11 Nov '08

11 Nov '08

Hello Max, Tuesday, November 11, 2008, 11:50:28 PM, you wrote: btw, i made here some time ago proposal about pure hashtables implented over a pure arrays (via accumArray operaion). may be it is somewhat helpful for you >>> > using unsafeFreeze. I'm getting stuck here, since the >>> IntMap library >>> > is not so monad-friendly. >>> >>> Data.Hashtable is >> >> Well, I need mutable update for a while... after that, I prefer a pure >> interface, which is why I'm trying to freeze all the values. > Is there any reason why an immutable Hashtable shouldn't be > implemented, with a pure interface, such that hashtables can be > (unsafe)Freezed into immutable hashtables? Would this be useful? > --Max -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

2 4

mapping unfreeze over an IntMap of IOUArrays
by Chad Scherrer 11 Nov '08

11 Nov '08

Hello cafe, I've hit a bit of a monadic snag here... I'm scanning a big file, building a table of statistics. I end up with something like IO (IntMap (IOUArray Int Double)) Once I've read in the whole file and built my statistics, I don't need any more updates, so I'd like to do something like IntMap (IOUArray Int Double) -> IO (IntMap (UArray Int Double)), using unsafeFreeze. I'm getting stuck here, since the IntMap library is not so monad-friendly. I could rebuild the whole thing using sequence and lists, but the table is pretty big (4.5 million keys on a first run), so I'd prefer to avoid that. Any ideas? BTW, I probably "should" be using ST for this, but I hit the usual "type s escapes" irritation and gave up. If that would work more easily, that would be fine with me too. Thanks! Chad Scherrer

5 6

Compilation issue with GHC 6.10.1 on Vista
by Alexey Romanov 11 Nov '08

11 Nov '08

I'm trying to update FreshLib <http://homepages.inf.ed.ac.uk/jcheney/programs/freshlib/> to work with the current GHC version. Now it works fine in GHCi, but trying to compile with ghc --make gives this result: [1 of 5] Compiling NominalBase ( NominalBase.hs, NominalBase.o ) Loading package ghc-prim ... linking ... done. Loading package integer ... linking ... done. Loading package base ... linking ... done. Loading package syb ... linking ... done. Loading package base-3.0.3.0 ... linking ... done. Loading package array-0.2.0.0 ... linking ... done. Loading package packedstring-0.1.0.1 ... linking ... done. Loading package containers-0.2.0.0 ... linking ... done. Loading package pretty-1.0.1.0 ... linking ... done. Loading package template-haskell ... linking ... done. Loading package bytestring-0.9.1.4 ... linking ... done. Loading package syb-with-class-0.4 ... linking ... done. C:\Users\Alexey\AppData\Local\Temp\/ghc1892_0/ghc1892_0.s: Assembler messages: C:\Users\Alexey\AppData\Local\Temp\/ghc1892_0/ghc1892_0.s:557: Error: symbol `_N ominalBase_zdgtoName_closure' is already defined C:\Users\Alexey\AppData\Local\Temp\/ghc1892_0/ghc1892_0.s:585: Error: symbol `_N ominalBase_zdgtoName_info' is already defined C:\Users\Alexey\AppData\Local\Temp\/ghc1892_0/ghc1892_0.s:637: Error: symbol `_N ominalBase_zdgfromName_closure' is already defined C:\Users\Alexey\AppData\Local\Temp\/ghc1892_0/ghc1892_0.s:665: Error: symbol `_N ominalBase_zdgfromName_info' is already defined Yours, Alexey Romanov

1 0

Problem installing curl
by Martin Huschenbett 11 Nov '08

11 Nov '08

Hi all, when I try to install curl (needed for hxt) using "cabal install curl" I alwas get the following error message: Resolving dependencies... 'curl-1.3.2.1' is cached. Configuring curl-1.3.2.1... cabal: Error: some packages failed to install: curl-1.3.2.1 failed during the configure step. The exception was: sh: runGenProcess: does not exist (No such file or directory) I don't know what to do here. Can anybody help me please? I'm using the brand new GHC 6.10.1 on Windows Vista. Best regards, Martin.

3 2

compilation question
by Peter Padawitz 11 Nov '08

11 Nov '08

At first a type of arithmetic expressions and its generic evaluator: data Expr = Con Int | Var String | Sum [Expr] | Prod [Expr] | Expr :- Expr | Int :* Expr | Expr :^ Int data ExprAlg a = ExprAlg {con :: Int -> a, var :: String -> a, sum_ :: [a] -> a, prod :: [a] -> a, sub :: a -> a -> a, scal :: Int -> a -> a, expo :: a -> Int -> a} eval :: ExprAlg a -> Expr -> a eval alg (Con i) = con alg i eval alg (Var x) = var alg x eval alg (Sum es) = sum_ alg (map (eval alg) es) eval alg (Prod es) = prod alg (map (eval alg) es) eval alg (e :- e') = sub alg (eval alg e) (eval alg e') eval alg (n :* e) = scal alg n (eval alg e) eval alg (e :^ n) = expo alg (eval alg e) n Secondly, a procedural version of eval (in fact based on continuations): data Id a = Id {out :: a} deriving Show instance Monad Id where (>>=) m = ($ out m); return = Id peval :: ExprAlg a -> Expr -> Id a peval alg (Con i) = return (con alg i) peval alg (Var x) = return (var alg x) peval alg (Sum es) = do as <- mapM (peval alg) es; return (sum_ alg as) peval alg (Prod es) = do as <- mapM (peval alg) es; return (prod alg as) peval alg (e :- e') = do a <- peval alg e; b <- peval alg e'; return (sub alg a b) peval alg (n :* e) = do a <- peval alg e; return (scal alg n a) peval alg (e :^ n) = do a <- peval alg e; return (expo alg a n) My question: Is peval less time- or space-consuming than eval? Or would ghc, hugs et al. optimize eval towards peval by themselves? Peter

3 2