- Haskell-Cafe - Haskell.org

Calling Haskell from Python / C++
by Jonathan Holt 14 Nov '02

14 Nov '02

Hi, I've just recently learned about Haskell, and I'm impressed by the abstractions and expressiveness that it affords. I'm particularly interested in it for a small parser project that I'm planning. However, my main programming languages are Python and C++, and for various reasons switching entirely to Haskell is completely out of the question (no flame-bait intended!). So here is what I envision: I write the main application in Python. I write a (hopefully) small Haskell module that: a) Calls back to the main Python app for reading the text to be parsed, preferably using laziness. b) Parses the text, and maybe processes a bit. c) Returns the parsed data-structure. (This may be tricky, but I think I know how to do it). This all seems quite cool. And it can also be very useful in a wide veriety of other circumstances, i.e. whenever a complex computation can be "outsourced" from Python/C++ to Haskell. But I can't find any way to interface Python and Haskell in this way. The FFI seems to allow for this, but there arn't any tools to actually do it... So my question is: How would one do such a thing? Thanks in advance, -JH __________________________________________________ Do you Yahoo!? U2 on LAUNCH - Exclusive greatest hits videos http://launch.yahoo.com/u2

3 2

Re: Time library underspecified
by Tom Pledger 14 Nov '02

14 Nov '02

John Meacham writes: : | another useful thing would be | endOfTime and beginningOfTime constants, representing the minimum and | maximum values representable by ClockTime. : | _______________________________________________ | Haskell mailing list | Haskell(a)haskell.org | http://www.haskell.org/mailman/listinfo/haskell If you plan to use endOfTime and beginningOfTime mainly as sentinels, and are happy to treat them as +infinity and -infinity for arithmetic, a separate wrapper type may help. I've found the following (abridged) useful for 'closing off' data types which were otherwise open-ended: > data Close a > = Lo > | Mid a > | Hi > deriving (Eq, Ord, Show) > > instance Bounded (Close a) where > minBound = Lo > maxBound = Hi > > instance Num a => Num (Close a) where ... > instance Real a => Real (Close a) where ... > instance Enum a => Enum (Close a) where ... > instance Integral a => Integral (Close a) where ... Regards, Tom

1 0

RE: Calling Haskell from Python / C++
by Simon Peyton-Jones 14 Nov '02

14 Nov '02

I hope it's all described adequately in the GHC user manual http://haskell.org/ghc If not, let us know! Simon | -----Original Message----- | From: Peter Simons [mailto:simons@cryp.to] | Sent: 13 November 2002 10:07 | To: haskell-cafe(a)haskell.org | Subject: Re: Calling Haskell from Python / C++ | | Simon Peyton-Jones writes: | | > If Python uses C's calling convention, it might be easy; just use | > 'foreign import' and 'foreign export' (see the FFI spec at | > haskell.org) | | I remember reading that the main program, that wants to call Haskell | code, would have to be compiled with a special Haskell main() routine, | though? Is that true? And if, why is that necessary? | | -peter | _______________________________________________ | Haskell-Cafe mailing list | Haskell-Cafe(a)haskell.org | http://www.haskell.org/mailman/listinfo/haskell-cafe

2 1

Library for symbolic integration/derivation/ODE solving?
by Marcus Lindblom 14 Nov '02

14 Nov '02

Hi! I would like a library that helps me with some symbolic integration/derivation/ODE solving. Sort of a Maple/Mathematica as a Haskell library, that operates on a expression tree. I've tried google and the haskell homepage library list but haven't found anything of the sort. A few pointers would be nice. ;) Regards /Marcus

1 0

RE: ghc binary, fptools CVS jamming :-(
by Simon Marlow 14 Nov '02

14 Nov '02

> Do I have to compile every packages again if I upgrade ghc ? Yes. > It is a pleasure to make use of the utilities (greencard, > HOpenGL, etc). But what > > When I try to use CVS repository and stand-alone binary distribution > together (because I wanted to see hdirect is like and v 0.16 > does not work with ghc-5.xx) I felt very much discouraged. > > What should I do ? > Do I always have to compile everything from the fptools source again ? If you install a new version of GHC, then you need to recompile anything which was compiled with the old version. If all you want to do is compile HOpenGL, say, then you don't need to install a new version of GHC. Cheers, Simon

1 0

HDirect
by Martin Huschenbett 14 Nov '02

14 Nov '02

Hi all together! Is there any newer version of HDirect than 0.17? Maybe an CVS-repository? THX, Huschi!

1 0

ghc binary, fptools CVS jamming :-(
by Ahn Ki-yung 14 Nov '02

14 Nov '02

Do I have to compile every packages again if I upgrade ghc ? It is a pleasure to make use of the utilities (greencard, HOpenGL, etc). But what When I try to use CVS repository and stand-alone binary distribution together (because I wanted to see hdirect is like and v 0.16 does not work with ghc-5.xx) I felt very much discouraged. What should I do ? Do I always have to compile everything from the fptools source again ? I hope FFI related Haskell module gets stable and standardized soon. -- Ahn Ki-yung

1 0

Re: Record of STRefs better than STRef to a Record?
by Jorge Adriano 14 Nov '02

14 Nov '02

> > I'm passing that 'record' around as an implicit value. The record as > > STRefs that I use to collect info, but it also has some pure fields with > > 'read-only' info. Something like, <opss let me change 'max' to 'm'> > > data E s = E{ > > refi :: STRef s Int, > > refc :: STRef s Char, > > m :: Int > > } > > > > In some functions I might need only some pure fields, and none of the > > STRefs, but since I pass something of type 'E s' around, everything ends > > up beeing monadic. > > there is no reason why that should be the case. you only need to be > within the ST monad when you read the references, so if in some function > you only look at the pure values, it can have a pure type. > hope this helped > iavor That's what I thought, untill I tried it. Well, actually you must be right since the pure field defines a pure (projection) function... Hmmm, ok, can someone explain this to me, data E s = E{ refi :: STRef s Int, refc :: STRef s Char, m :: Int } -- this is fine, obviously... pure :: E s -> Int pure e = m e -- but this is not... pure2 :: (?e :: E s) => Int pure2 = m (?e) Why exactly isn't this allowed? What is the workaround? Error msg: ------------------------------------------------------------------------ Ambiguous constraint `?e :: E s' At least one of the forall'd type variables mentioned by the constraint must be reachable from the type after the '=>' In the type: forall s. (?e :: E s) => Int While checking the type signature for `pure2' Failed, modules loaded: none. -------------------------------------------------------------------------- Thanks, J.A.

1 0

Record of STRefs better than STRef to a Record?
by Jorge Adriano 13 Nov '02

13 Nov '02

Hi all, If I use an STRef to a record, will a new record be created each time I want to update a single field? Or can I expect GHC to optimize it and have the field of the record updated in place? Right now I'm using a record of STRefs, like: data E s = E{ refi :: STRef s Int, refc :: STRef s Char } but it can get a little messy, since thread s propagates to every datatype that uses the record type in it's definition. Thanks, J.A.

2 2

monadic stack to register machine translator
by William Lee Irwin III 13 Nov '02

13 Nov '02

module GT where import Monad import Monoid import MonadState import MonadWriter import MonadRWS -- Just a quick exercise in using monads. -- Thought it'd be nice to share with the class. data GOp = PushVal Integer | Push Integer | Pop Integer | Slide Integer | Update Integer | GAdd | GSub | GMul | GDiv | GMod | GPow | GNeg | GAbs deriving (Eq, Ord, Read, Show) type Tmp = Integer data ROp = LoadImm Tmp Integer | RAdd Tmp Tmp Tmp | RSub Tmp Tmp Tmp | RMul Tmp Tmp Tmp | RDiv Tmp Tmp Tmp | RMod Tmp Tmp Tmp | RPow Tmp Tmp Tmp | RNeg Tmp Tmp | RAbs Tmp Tmp deriving (Eq, Ord, Read, Show) type CounterT m t = StateT Integer m t type StackT t = State [Integer] t type GST t = RWS () [ROp] (Integer, [Integer]) t class Stack f where pushVal, push, pop, update, slide :: Integral t => t -> f () popVal :: Integral t => f t instance Integral t => Stack (RWS () [ROp] (t, [t])) where pushVal n = do (ctr, stk) <- get put (ctr, fromIntegral n : stk) popVal = do (ctr, top:stk) <- get put (ctr, stk) return (fromIntegral top) push n = do (ctr, stk) <- get put (ctr, stk!!fromIntegral n : stk) pop n = do (ctr, stk) <- get put (ctr, drop (fromIntegral n) stk) slide n = do (ctr, top:stk) <- get put (ctr, top : drop (fromIntegral n) stk) update n = do (ctr, top:stk) <- get let (front, _:back) = splitAt (fromIntegral n) stk put (ctr, front ++ [top] ++ back) class Counter f where gen :: Enum t => f t instance Integral t => Counter (RWS () [ROp] (t, [t])) where gen = do (ctr, stk) <- get put (ctr + 1, stk) return . toEnum . fromIntegral $ ctr + 1 instance (Enum t, Monad m) => Counter (StateT t m) where gen = do ctr <- get put $ succ ctr ctr <- get return . toEnum $ fromEnum ctr translate gOps = snd $ evalRWS (mapM trans gOps) () (0,[]) trans :: GOp -> GST () trans i = case i of PushVal n -> do reg <- gen tell [LoadImm reg n] pushVal reg Push n -> push n Pop n -> pop n Slide n -> slide n Update n -> update n GAdd -> doBinOp RAdd GSub -> doBinOp RSub GMul -> doBinOp RMul GDiv -> doBinOp RDiv GMod -> doBinOp RMod GPow -> doBinOp RPow GNeg -> doUnOp RNeg GAbs -> doUnOp RAbs where doUnOp op = do x <- popVal y <- gen tell [op y x] pushVal y doBinOp op = do x <- popVal y <- popVal z <- gen tell [op z x y] pushVal z

1 0