November 2005 - Haskell-Cafe

Re: [Haskell] specification of sum
by ajb＠spamcop.net 04 Nov '05

04 Nov '05

(Moving this to the cafe.) G'day all. Quoting Cale Gibbard <cgibbard(a)gmail.com>: > We already do rely on them in most cases. Of course, not every > property can be proved by the compiler, but many pieces of code are > going to assume quite a lot. Agreed. > I think that the assumption that (+) and (*) in Num define something > like a ring on the given type is a sensible one. I'm not so certain. Octonian multiplication, to pick one example, is not associative, but I'd like to be able to use (*) nonetheless. Cheers, Andrew Bromage

2 1

Fixup too large
by Joel Reymont 04 Nov '05

04 Nov '05

What does this mean exactly? ghc -prof --make -auto-all -o logon -package Script Logon.hs Chasing modules from: Logon.hs Compiling Main ( Logon.hs, Logon.o ) /tmp/ghc20361.s:17881:Fixup of 93184 too large for field width of 16 bits /tmp/ghc20361.s:17878:Fixup of 93196 too large for field width of 16 bits /tmp/ghc20361.s:17874:Fixup of 93160 too large for field width of 16 bits -- http://wagerlabs.com/

1 1

Catching an escaping exception
by Joel Reymont 04 Nov '05

04 Nov '05

How can I catch this exception and translate it into an error within my monad? The exception is going to ghci and is not being caught in the monad. I'm interested in catching it and failing the monad gracefully. Exception: <interactive>: connect: does not exist (Connection refused) Types follow... type EngineState = ErrorT String (StateT World IO) type EngineResult = IO (Either String (), World) connect :: [Prop] -> HostName -> Int -> EngineState () connect env h p = do w <- get let secs = timeout_seconds w (Right h) <- liftIO $ timeout secs $ connect_ h p trace "Connection established" ... connect_ :: HostName -> Int -> IO Handle connect_ h p = connectTo h $ PortNumber $ fromIntegral p timeout :: forall a.Int -> IO a -> IO (Either String a) timeout seconds action = ... Thanks, Joel -- http://wagerlabs.com/

2 1

RE: [Haskell-cafe] handling rank 2 types
by Ralf Lammel 04 Nov '05

04 Nov '05

In a type system like MLF, your original code may type check. Let's do an experiment. We replace the IO monad by the Id(entity) monad. We actually replace the Id newtype also to become true type-level identity. So we get: -- -- This is like your 2nd say unchecked... sample -- fooBar :: [v] -> Empty fooBar l = Empty (map no l) where no :: a -> a' no x = error "element found" But the problem is not about "a higher-rank type occurring with a type constructor", as you guess. It is rather about functions with classic rank-1 types. To see this, insert $ for the application of Empty: fooBar :: [v] -> Empty fooBar l = Empty $ (map no l) where no :: a -> a' no x = error "element found" You get the same type error as for the original monadic code. What you could do is define a suitably typed application operator (likewise, a suitably typed liftM). (With MLF the various types would be admitted by a more general type.) In the non-monadic example, we need: -- Use app instead of ($) app :: ((forall a. [a]) -> c) -> (forall b. [b]) -> c app f x = f x BTW, what operations are we supposed to perform on the content of Empty; what's the real purpose for introducing this independent type variable "a'"? Parametric polymorphism seems to suggest that we can't do much. You can observe the length of the list. That's it, more or less. Why not store the length directly rather than having this superficially polymorphic list? I am a little bit lost. You wonder how you can generally solve this sort of problem? - You seem to spot the wrapping approach: wrap foralls and unpack at application time. This makes particularly sense for polymorphic *functions*. - The other approach is to use consistently rank-2 combinators. I probably don't get what you try to do. Ralf > -----Original Message----- > From: haskell-cafe-bounces(a)haskell.org [mailto:haskell-cafe- > bounces(a)haskell.org] On Behalf Of Andrew Pimlott > Sent: Thursday, November 03, 2005 8:35 PM > To: haskell-cafe(a)haskell.org > Subject: [Haskell-cafe] handling rank 2 types > > I am trying to use a rank 2 type to enforce a condition on a data > structure. For the purpose of this message, I am trying to ensure > that a list is empty (ignoring the possibility of bottom elements): > > > {-# OPTIONS -fglasgow-exts #-} > > import Control.Monad > > newtype Empty = Empty (forall a. [a]) > > I want a function > > > listToEmpty :: [v] -> IO Empty > > listToEmpty l = liftM Empty (mapM no l) where > > no :: a -> IO a' > > no x = fail "element found" > > But this gives a "less polymorphic" error. I think the problem is > that ghc will not instantiate the "a'" in "IO a'" as a higher-rank > type because it occurs within a type contructor. I believe this is > the restriction referred to at the end of 7.4.9[1]. If I instead > write > > > uncheckedListToEmpty :: [v] -> Empty uncheckedListToEmpty l = Empty > > (map no l) where > > no :: a -> a' > > no x = error "element found" > > it compiles, because "v'" can be instantiated as "forall v. Rule v", > and even pass through map with the higher-rank type. > > Is there any way to make ghc accept the first definition? I found > this > workaround: > > > newtype Any = Any { unAny :: forall a. a } listToEmptyAny :: [v] -> > > IO Empty listToEmptyAny l = liftM (\l -> Empty (map unAny l)) (mapM > > no l) where > > no :: a -> IO Any > > no x = fail "element found" > > But needing a newtype wrapper Empty was bad enough; do I really need > one for every intermediate result (that is inside another type > constructor) as well? I could probably define a generic family of > wrappers > > > newtype Forall = Forall (forall a. a) newtype Forall1 c1 = Forall1 > > (forall a. c1 a) newtype Forall2 c1 c2 = Forall2 (forall a. c1 c2 a) > ... > > But I would still have to use them everywhere. Any other tricks? > > Andrew > > [1] http://haskell.org/ghc/docs/latest/html/users_guide/type- > extensions.html#universal-quantification > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe

2 1

RE: [Haskell-cafe] handling rank 2 types
by Simon Peyton-Jones 04 Nov '05

04 Nov '05

| But this gives a "less polymorphic" error. I think the problem is that | ghc will not instantiate the "a'" in "IO a'" as a higher-rank type | because it occurs within a type contructor. I believe this is the | restriction referred to at the end of 7.4.9[1]. If I instead write I think that is indeed the heart of your problem. Actually I'm writing a paper right now about how to lift this restriction. An out of date version is here, http://research.microsoft.com/%7Esimonpj/papers/boxy/ that gives the general idea, but a much better version will be done in a week's time. It's not fully implemented yet, but I'll keep your message as an example to try when it is. Simon -cafe

1 0

newbie help
by Abhijit Ray 04 Nov '05

04 Nov '05

Hi I just started on haskell , i am using the yet another haskell tutorial by Hal Daume I wrote the following program and it dint compile. what's wrong with it module Main where import IO main = do hSetBuffering stdin LineBuffering let testList = makeList let sum = foldr (+) 0 testList putStrLn "Sum is " ++ show(sum) makeList = do putStrLn "enter a num" num <- getLine let nbr = read num if nbr == 0 then do return [] else do all <- makeList return (nbr : all) ---------------------------------------------- the error i get is Chasing modules from: Sum.hs Compiling Main ( Sum.hs, Sum.o ) Sum.hs:6: Couldn't match `[a]' against `IO ()' Expected type: [a] Inferred type: IO () In the application `putStrLn "Sum is "' In the first argument of `(++)', namely `putStrLn "Sum is "' Sum.hs:9: Couldn't match `[b]' against `IO [a]' Expected type: [b] Inferred type: IO [a] In the third argument of `foldr', namely `testList' In the definition of `sum': sum = foldr (+) 0 testList ----------------------------- Thanks, Abhijit Ray

3 3

Threads talking back to parent
by Joel Reymont 03 Nov '05

03 Nov '05

Folks, I'm launching thousands of threads that send and receive commands. I would like the launching thread to keep a log of all the commands sent and received, in proper order. This would mean that each time a command is sent and received in the child thread it would need to be sent back to parent who would log the command on a first-come first-serve basis. How would I implement this in Haskell? Thanks, Joel -- http://wagerlabs.com/

3 4

Maximum number of sockets
by Joel Reymont 03 Nov '05

03 Nov '05

Is there a limit on the maximum number of sockets that's possible to open in GHC? I'm concerned with Mac OSX and Windows. Thanks, Joel -- http://wagerlabs.com/

1 0

Socket finalizers
by Joel Reymont 03 Nov '05

03 Nov '05

Folks, I remember a message about this recently and tried googling for it but came up empty. I'm running a bunch of "poker bots" from ghci and although I'm calling hClose on the socket handle, it does not seem to actually close when my bots finish. I need to exit ghci for the sockets to close. How can I tackle this problem? It's important for me that the sockets close when I call hClose . Thanks, Joel -- http://wagerlabs.com/

2 1

Compiled hugs - on "overload" crashes
by Dusan Kolar 03 Nov '05

03 Nov '05

Hello again, without the --enable-timer option, I have managed to compile hugs. Nevertheless, I was not successful with doc. OK, I don't care about it, but anyway... I've run several tests. One of them is appended - for trial just type: cesty (1,1) /(It's a kind of state-space search implementation (horse problem) made by a student long time ago, without any optimization, for testing very good.)/ My older hugs interpreter stops with error: ERROR - Control stack overflow GHCi stops with the same error: *** Exception: stack overflow That's OK, really, that's correct behavior after printing some results. Nevertheless, the new hugs compilation stops with this: Segmentation Fault(coredump) And that's probably not the correct end. May I do something wrong? I've thought that configure + make + make install_all_but_docs (no doc compiled) should work. Sorry for bothering if this is trivial or off topic. Regards, hoping for help, Dusan -- Dusan Kolar tel: +420 54 114 1238 UIFS FIT VUT Brno fax: +420 54 114 1270 Bozetechova 2 e-mail: kolar(a)fit.vutbr.cz Brno 612 66 Czech Republic --

1 0