- Haskell-Cafe - Haskell.org

Request for subject suggestions for a series of informal top 10 Haskell articles
by Kenneth Hoste 09 Nov '05

09 Nov '05

Greetings, As you all might know, the Haskell community has it's own informal magazine, The Monad.Reader (http://haskell.org/tmrwiki) I'm planning to write (with some support from the frequent users of #haskell @ irc.freenode.net) a series of articles for TMR, which describe some aspect of Haskell. Because of the informal (but serious and thorough) style of TMR, I'd like to do this using a top 10 list regarding some subject. Every article will list 10 of most famous, most frequently used or most usefull things regarding that subject. I'd like to ask for suggestions on each subject when I'm writing the article on this mailing list, because I don't know the Haskell world very well (I've only been working with it for a year know, in my spare time). This first mail is intented to ask for suggestions regarding future subjects for the articles. Suggestions are welcome by replying, or by adding them to the 'take 10 [1..]'-wiki at http://haskell.org/tmrwiki/TopTenHaskell . Also other suggestions or remarks are very welcome. I hope this series of articles can provide a source of information for both new and experienced users, because you there's always room to discover new aspects of your favourite language. Kenneth a.k.a. boegel

1 0

Inferred type less polymorphic than expected
by Joel Reymont 09 Nov '05

09 Nov '05

Folks, How do I fix this? data Prop = forall a b.(Eq a, Show a) => Attr a := a data Attr a = Attr String (a -> Dynamic, Dynamic -> Maybe a) (PU a) type Props = M.Map String (Int, Prop) instance Ord (Int, Prop) where compare (a, _) (b, _) | a == b = EQ | a > b = GT | otherwise = LT makeAttr :: Typeable a => String -> PU a -> Attr a makeAttr name pickler = Attr name (toDyn, fromDynamic) pickler ... props :: Props -> PU Props props m = props' $ sort $ M.toList m where props' [] = lift [] props' ((_, (Attr _ _ pp := _)):xs) = wrap (\(a, b) -> a : b, \(a : b) -> (a, b)) (pair pp (props' xs)) ./Script/Prop.hs:80:10: Inferred type is less polymorphic than expected Quantified type variable `a' is mentioned in the environment: props' :: [(a1, Prop)] -> PU [a] (bound at ./Script/Prop.hs: 79:10) When checking an existential match that binds $dEq :: {Eq a} $dShow :: {Show a} pp :: PU a The pattern(s) have type(s): [(a1, Prop)] The body has type: PU [a] In the definition of `props'': props' ((_, (Attr _ _ pp := _)) : xs) = wrap (\ (a, b) -> a : b, \ (a : b) -> (a, b)) (pair pp (props' xs)) In the definition of `props': props m = props' $ (sort $ (Data.Map.toList m)) where props' [] = lift [] props' ((_, (Attr _ _ pp := _)) : xs) = wrap (\ (a, b) -> a : b, \ (a : b) -> (a, b)) (pair pp (props' xs)) Thanks, Joel -- http://wagerlabs.com/

2 2

Re: [Haskell-cafe] Inferred type less polymorphic than expected
by Joel Reymont 09 Nov '05

09 Nov '05

This gets me pretty close. I now have to figure out how to return a PU Props instead of a PU [Prop] which is what it's returning. Still, it's a huge step forward. props :: Props -> PU Props props m = props' $ sort $ M.toList m where props' [] = lift [] props' ((_, (_, x)):xs) = wrap (\(a, b) -> a : b, \(a : b) -> (a, b)) (pair (prop x) (props' xs)) ./Script/Prop.hs:82:10: Couldn't match `Props' against `[Prop]' Expected type: PU Props Inferred type: PU [Prop] In the expression: props' $ (sort $ (Data.Map.toList m)) In the definition of `props': props m = props' $ (sort $ (Data.Map.toList m)) where props' [] = lift [] props' ((_, (_, x)) : xs) = wrap (\ (a, b) -> a : b, \ (a : b) -> (a, b)) (pair (prop x) (props' xs)) Big thanks to Andrew and Malcolm on #haskell for pointing me in the right direction! Thanks, Joel On Nov 8, 2005, at 5:23 PM, Andrew Pimlott wrote: > You might have better luck getting if you send a complete example, > since > it's hard for me to figure out exactly what this code is supposed > to do. > However, I'll take a quick stab. > > On Tue, Nov 08, 2005 at 04:46:56PM +0000, Joel Reymont wrote: >> props :: Props -> PU Props >> props m = props' $ sort $ M.toList m >> where props' [] = lift [] >> props' ((_, (Attr _ _ pp := _)):xs) = >> wrap (\(a, b) -> a : b, >> \(a : b) -> (a, b)) >> (pair pp (props' xs)) >> >> >> ./Script/Prop.hs:80:10: >> Inferred type is less polymorphic than expected >> Quantified type variable `a' is mentioned in the environment: >> props' :: [(a1, Prop)] -> PU [a] (bound at ./Script/Prop.hs: >> 79:10) > > Recall that pp is a (PU a) for some a. However, the inferred type for > props' is fully polymorphic in a. It might help first of all to write > down the type you expect for props', and then add type annotations to > different pieces until you find the one that's going awry. > > Andrew -- http://wagerlabs.com/

1 0

\Parsec\Token.hs
by Sara Kenedy 08 Nov '05

08 Nov '05

Dear all, I run file Token.hs in hugs98\libraries\Text\ParserCombinator\Parsec\Token.hs, but it displays this error ERROR : 64 - Syntax error in data type definition (unexpected '.') Anyone know how to fix this? Thanks.

4 3

RE: [Haskell-cafe] Type classes and hFoldr from HList
by Ralf Lammel 08 Nov '05

08 Nov '05

Hi Greg, Since hfoldr is right-associative, I prefer to reorder your list of functions as follows: > test = HCons (length::String -> Int) (HCons ((+1)::(Int->Int)) (HCons ((*2)::(Int->Int)) HNil)) Note that I also annotated length with its specific type. (If you really wanted to leave things more polymorphic, you would need to engage in TypeCast.) Providing a specific Apply instance for (.) is not necessary, strictly necessary. We could try to exploit the normal function instance for Apply. Let me recall that one here for convenience: >instance Apply (x -> y) x y > where > apply f x = f x Let me also recall the hFoldr instances: >class HList l => HFoldr f v l r | f v l -> r > where > hFoldr :: f -> v -> l -> r >instance HFoldr f v HNil v > where > hFoldr _ v _ = v >instance ( HFoldr f v l r > , Apply f (e,r) r' > ) > => HFoldr f v (HCons e l) r' > where > hFoldr f v (HCons e l) = apply f (e,hFoldr f v l) To fit in (.), we would flip and uncurry it. So we could try: comp' = hFoldr (uncurry (flip (.))) (id::Int -> Int) test This wouldn't work. The trouble is the required polymorphism of the first argument of hFoldr. The type of that argument as such is polymorphic. However, this polymorphism does not survive type class parameterization. You see this by looking at the HCons instance of HFoldr. The different occurrences of "f" would need to be used at different types. This would only work if the type class parameter f were instantiated by the polymorphic type of (uncurry (flip (.))). (And even then we may need something like TypeCast.) What you can do is define a dedicated *type code* for composition. comp = hFoldr (undefined::Comp) (id::Int -> Int) test data Comp instance Apply Comp (x -> y,y -> z) (x -> z) where apply _ (f,g) = g . f Ralf > -----Original Message----- > From: haskell-cafe-bounces(a)haskell.org [mailto:haskell-cafe- > bounces(a)haskell.org] On Behalf Of Greg Buchholz > Sent: Sunday, November 06, 2005 7:01 PM > To: haskell-cafe(a)haskell.org > Subject: [Haskell-cafe] Type classes and hFoldr from HList > > > I was playing around with the HList library from the paper... > > Strongly typed heterogeneous collections > http://homepages.cwi.nl/~ralf/HList/ > > ...and I thought I'd try to fold the composition function (.) through a > heterogeneous list of functions, using hFoldr... > > >{-# OPTIONS -fglasgow-exts #-} > >{-# OPTIONS -fallow-undecidable-instances #-} > > > >import CommonMain > > > >main = print $ comp "abc" > > > >test = HCons ((+1)::(Int->Int)) (HCons ((*2)::(Int->Int)) (HCons length > HNil)) > > > >comp = hFoldr (.) id test > > > >instance Apply (a -> b -> c -> d) (a, b) (c -> d) > > where > > apply f (a,b) = f a b > > ...but it fails with the following type error... > > ]Compiling Main ( compose.hs, interpreted ) > ] > ]compose.hs:10:7: > ] No instances for (Apply ((b -> c) -> (a -> b) -> a -> c) > ] (Int -> Int, r) > ] ([Char] -> a3), > ] Apply ((b -> c) -> (a -> b) -> a -> c) (Int -> Int, > r1) r, > ] Apply ((b -> c) -> (a -> b) -> a -> c) ([a2] -> > Int, a1 ->a1) r1) > ] arising from use of `hFoldr' at compose.hs:10:7-12 > ] Probable fix: > ] add an instance declaration for (Apply ((b -> c) -> (a -> b) -> a - > > c) > ] (Int -> Int, r) > ] ([Char] -> a3), > ] Apply ((b -> c) -> (a -> b) -> a - > > c) > ](Int -> Int, r1) r, > ] Apply ((b -> c) -> (a -> b) -> a - > > c) > ]([a2] -> Int, a1 -> a1) r1) > ] In the definition of `comp': comp = hFoldr (.) id test > > ...Anyway, I couldn't quite tell whether I was using hFoldr incorrectly, > or if I needed to have more constraints placed on the construction of > "test", or if needed some sort of type-level function that resolves... > > Apply ((b -> c) -> (a -> b) -> a -> c) > > ...into (a -> c), or something else altogether. I figured someone might > be able to help point me in the right direction.

2 1

Information on B-Tree IO implemenations
by Scott Weeks 08 Nov '05

08 Nov '05

Hello all, (I'm resending this because I sent the first one without being a member of the list and I'm assuming it bounced) I'm a newbie and while I grasp some of the basics of FP I have yet to really understand the guts of Haskell. The type system is starting to clear up for me and so are Monads but I have a long way to go. For a project that I'm working on I need to implement a B-Tree or similar disk based data structure for storing large amounts of data. I feel like I'm getting it and then I get hung up on disk IO operations and trying to figure out how they should fit into the mix. I'm not necessarily looking for someone to point me at an implementation (although I wouldn't mind) but if someone could point me in the direction of some material that could shed a light on things I would be much obliged. I've read some Okasaki and found a promising paper called "Database Manipulation in Haskell 1.3" by Kevin Hammond and Phil Trinder but while they talk to the data structures themselves they don't quite get around to mentioning how they achieve the IO part of the equation. Thanks, Scotty Weeks

2 2

RE: Re[2]: [Haskell-cafe] Re: FPS: Finalizers not running (wasMemoryusageoutside of the Haskell heap)
by Simon Marlow 08 Nov '05

08 Nov '05

On 07 November 2005 14:02, Bulat Ziganshin wrote: > Hello Simon, > > Monday, November 07, 2005, 4:16:54 PM, you wrote: > >> Yes, mallocForeignPtr's are implemented using pinned GC'd memory, >> with no finalizer (unless you add one). That's why they're nice and >> fast. > > Simon, can you please write some docs about all these foreign stuff > and their interaction with pure Haskell code? i see some bits of this > information in these newsgroups regularly (for example, are you > remember discussion about using lots of hashtables in program?), but > there is no definitive source > > there is no need to write full-fledged doc right now, just > "implementation notes" as the list of peculiarities will be great for > the start You're right, it would be good to have this folklore stuff written down. I've added some notes to the mallocForeignPtr docs for now. Cheers, Simon

1 0

Monadifying pickler combinators
by Joel Reymont 08 Nov '05

08 Nov '05

Folks, I'm trying to monadify http://research.microsoft.com/~akenn/fun/ picklercombinators.pdf appU is easy but I can't figure out the appP. Is it actually possible? The clever trick seems to be the passing of (b -> a) into sequ for use in appP. -- instance Monad (PU s) where return a = PU snd (\s -> (a, s)) -- how do I do appP below? pa >>= k = PU (\(b, s) -> let pb = k a in appP pa (a, ??? -- this is good (\s -> let (a, s1) = appU pa s pb = k a in appU pb s1) Thanks, Joel -- http://wagerlabs.com/

1 0

RE: [Haskell-cafe] Trashing of memory: How to troubleshoot and fix?
by Simon Marlow 08 Nov '05

08 Nov '05

On 07 November 2005 14:16, Joel Reymont wrote: > Well, I was suspecting FPS but I removed all FPS usage and went with > regular lists. The program still crashes. I did not see a "malloc: > memory exhausted" error from ghci this time, it just crashed with an > application error while writing to some random memory location. > > I built a separate exe (this is Windows) and ran it in gdb for a > segmentation violation in with this stack trace: > > ?? () > s7yw_info () > ?? () If you are pretty certain that you aren't doing anything unsafe, then this is a bug in GHCi. If you can provide us with a reproducible test case, I'll look into it. Cheers, Simon

1 0

RE: [Haskell-cafe] Trashing of memory: How to troubleshoot and fix?
by Simon Marlow 07 Nov '05

07 Nov '05

On 05 November 2005 10:10, Joel Reymont wrote: > This is a I came up with a "scripting" environment for poker server > people to excercise their server. All it basically does is compose > and parse binary packets and let scripters send and receive them. > > The issue I have is that the "script" runs fine on Mac OSX (1Gb of > memory, though 6Gb VM) and runs out of memory on a Windows machine > with 512Mb of memory and 768Mb of VM. The memory error is from malloc > which leads me to believe memory fragmentation could be an issue. > Running on Windows is a customer requirement. GHC very rarely calls malloc(), so this is likely to be caused by heavy memory use in the C++ part of your app. It would be interesting to find out though: you can run gdb on the binary and get a backtrace when the memory error occurs (compiling the Haskell app with -debug will help). > I was wondering if peak memory usage was the issue but jugding from > the GC stats it does not seem to be that high. Maybe I'm wrong. What > do you guys make of this? > > ./logon +RTS -p -sblah > > GC summary: > -- > 5,695,379,620 bytes allocated in the heap > 1,274,038,800 bytes copied during GC > 69,790,544 bytes maximum residency (53 sample(s)) > > 20433 collections in generation 0 (230.29s) > 53 collections in generation 1 ( 5.15s) > > 152 Mb total memory in use > > INIT time 0.00s ( 0.04s elapsed) > MUT time 91.13s (250.79s elapsed) > GC time 235.44s (607.31s elapsed) > RP time 0.00s ( 0.00s elapsed) > PROF time 0.00s ( 0.00s elapsed) > EXIT time 0.00s ( 0.00s elapsed) > Total time 326.57s (858.14s elapsed) GC load is *very* high. But as you can see, GHC's storage manager is only using 152Mb peak - if you're running out with 512Mb, then something else is using the rest. To improve performance, I would first try to optimise the program to reduce the allocation rate, get any low-hanging fruit, and then tweak the GC settings. Most of your GC is generation 0 collections, which suggests that your program is allocating a lot of temporary memory and not getting much work done. Also, note that you got these figures from a version of the program compiled for profiling, which will have higher memory requirements than normal due to the profiling overheads. > --- > Time allocation and profiling report: > ---- > total time = 85.58 secs (4279 ticks @ 20 ms) > total alloc = 2,941,143,704 bytes (excludes profiling > overheads) > > COST CENTRE MODULE %time %alloc > > exp/evalU/cmd Script.Engine 50.0 66.4 > connect Main 14.1 11.1 > CAF Script.Engine 10.3 5.4 > exp/if Script.Engine 8.8 4.8 > send Main 6.4 6.7 > expect Main 5.8 2.7 > exp/evalU/sz Script.Engine 1.4 1.4 > startSSL Script.Engine 1.3 0.1 > --- Seems like you need to look at evalU: > cmd' <- {-# SCC "exp/evalU/cmd" #-}return $ evalU packet appU > -- unpickle command this is where most of the time/alloc seems to be going. Also, just looking at this line of code reminds me of SimonM's 203rd law of Haskell optimisation: when writing 'return $ E', consider writing 'return $! E' instead. Cheers, Simon

2 1