- Haskell-Cafe - Haskell.org

Another fold question
by Patty Fong 07 Nov '03

07 Nov '03

Having struglled over this for the better part of a day and only becoming more frustrated the more i try to understand it, i once again seek help :) I understand how basic folds work, i.e foldr replaces (:) with some parameter and [] by another i.e foldr (+) 0 [1,2,3] becomes 1+(2+(3+0)) I also understand how to write my own fold function. What i don't understand quite is how to use them. given this data type and this fold function i wrote: data Music = Note Pitch Octave Duration | Silence Duration | PlayerPar Music Music | PlayerSeq Music Music | Tempo (Ratio Int) Music data Pitch = Cf | C | Cs type Octave = Int type Duration = Ratio Int foldMusic :: (Pitch -> Octave -> Duration -> a) -> (Duration -> a) -> (a -> a -> a) -> (a -> a -> a) -> (Ratio Int -> a -> a) -> Music -> a foldMusic n _ _ _ _ (Note pitch octive duration) = n pitch octive duration foldMusic _ s _ _ _ (Silence duration) = s duration foldMusic n s p1 p2 t (PlayerPar partOne partTwo) = p1 (foldMusic n s p1 p2 t partOne)(foldMusic n s p1 p2 t partTwo) foldMusic n s p1 p2 t (PlayerPar partA partB) = p2 (foldMusic n s p1 p2 t partA)(foldMusic n s p1 p2 t partB) foldMusic n s p1 p2 t (Tempo rate part) = t rate (foldMusic n s p1 p2 t part) I understand that when i use the foldMusic function i need to pass it 5 parameters. given the type signiature, why can i pass (+) as a parameter for p1 but not for n, what determines what can be passed as a parameter, because they all have the return type a?? I attempted to create a function that utilises the foldMusic function that counts the number of notes: count_notes :: Music -> Integer count_notes = foldMusic (\_-> \_ -> \_ -> 1) (\_ -> 0) (+) (+) (\_ -> \_ -> 0) it appears to work, i think. Yet i'm still not certain of how it does so. This confuses me, Is there anyway to represent other fold functions in a tree like representation as foldr (+) 0 would appear as such? + 1 \ + 2 \ + 3 \ 0 Regards, confused Patrick _________________________________________________________________ Hot chart ringtones and polyphonics. Go to http://ninemsn.com.au/mobilemania/default.asp

9 10

Another Newbie question :)
by Patty Fong 06 Nov '03

06 Nov '03

I seem to understand the basic idea behind the fold function, i think but do not seem to be able to write them myself. My understanding is that it works by replacing every constructor with a fold ? Could anyone please point me in the direction of a suitable resource of attempt to explain this to me :) Cheers, Paty _________________________________________________________________ Hot chart ringtones and polyphonics. Go to http://ninemsn.com.au/mobilemania/default.asp

4 4

Trying to understand tuples and lists
by Karthik Kumar 05 Nov '03

05 Nov '03

Hi, I have a list of tuples ( characters and numbers ) . say, [('a', 1), ('b', 2)] I want to extract 'a' and 1 separately . (x,y) <- head [('a', 1), ('b', 2)] This doesnt work but my idea is to get x contain the value 'a' and y contain the value 1 as the first tuple in the list. Which is the right way of doing this. Thanks for your help. Cheers Karthik. __________________________________ Do you Yahoo!? Protect your identity with Yahoo! Mail AddressGuard http://antispam.yahoo.com/whatsnewfree

3 3

Type errors in Haskell programming languages - Plz help
by Karthik Kumar 05 Nov '03

05 Nov '03

Hi, I was trying this Haskell program and was getting the following type error. The code is as follows - <-- Code starts --> entry :: [Char] -> [(Char,Int)] entry list = do t <- getGroups list mergeGroups t getGroups :: [Char] -> [(Char,Int)] mergeGroups :: [(Char,Int)] -> [(Char,Int)] <-- Code Ends --> The function getGroups and mergeGroups compile perfectly fine. The error i am getting is in the function entry . To paste the message, " *** Expression : mergeGroups t *** Term : t *** Type : (Char,Int) *** Does not match : [(Char,Int)] " Thanks in advance for helping me in this regard. Cheers Karthik. __________________________________ Do you Yahoo!? Protect your identity with Yahoo! Mail AddressGuard http://antispam.yahoo.com/whatsnewfree

2 2

RE: Type tree traversals [Re: Modeling multiple inheritance]
by Simon Peyton-Jones 05 Nov '03

05 Nov '03

| We really should change GHC rather than keep trying to work around stuff | like this. GHC will be my light reading for winter break. Maybe so. For the benefit of those of us who have not followed the details of your work, could you summarise, as precisely as possible, just what language extension you propose, and how it would be useful? A kind of free-standing summary, not assuming the reader has already read the earlier messages. Simon

3 2

Re: Data types basics
by Patty Fong 05 Nov '03

05 Nov '03

Another prompt reply! thanks Hal :) I think I understand this correctly now For my previous problem which was: <prolog> ::= (<assertion> ".")* <assertion> :: = <structure> | <rule> <rule> ::= <structure> ":-" <structure>("," <structure>)* <structure> ::= <name> [( <term> (, <term>)* )] <term> ::= <number> | <variable> | <structure> <variable> ::= <name> Does this appear to be correct: type Prolog = [Assertion] data Assertion = StrucAssert Structure | RuleAssert Rule data Rule = Rule Structure Structure [Structure] data Structure = Structure Name Term [Term] data Term = NumTerm Number | VarTerm Variable | StrucTerm Structure type Variable = Name type Name = String type Number = Int >From: Hal Daume III To: Patty Fong CC: Haskell Cafe , Hal Daume Subject: >Re: Data types basics Date: Tue, 4 Nov 2003 22:10:03 -0800 (PST) > >Hi again, > >On Wed, 5 Nov 2003, Patty Fong wrote: > > > Hi to anyone reading this. i'm still strugling a bit with data type > >declarations. > > The was i understand it is that if i delcare a new data >type: > > data myType = myType a | b | c > >This isn't entirely correct. The names of types have to begin with capital >letters, as do constructors (more later). So you would need this to be: > >data MyType = MyType A | B | C > >where A is an existing type. > >This type now has three constructors: > >MyType :: A -> MyType B :: MyType C :: MyType > >It's perhaps a bit easier to understand when the names are different. > >When we say: > >data Foo = Bar Int | Baz String > >this means that a "Foo" is of one of two forms (the "|" can be read as >disjunction). A value of type Foo is either of the form "Bar x" for some x >which is an Int or "Baz y" for some y which is a String. > >"Bar" and "Baz" are called "constructors" because they take arguments and >"construct" a Foo. So, in this case, > >Bar :: Int -> Foo Baz :: String -> Foo > >are the two constructors. > >Of course, you have have any number of constructors and each can have any >number of arguments. > >data Foo = Bar | Baz Int | Bazaa String Bool > >Now there are three constructors: > >Bar :: Foo Baz :: Int -> Foo Bazaa :: String -> Bool -> Foo > >they can be recursive: > >data Foo = Bar | Baz Int Foo > >Bar :: Foo Baz :: Int -> Foo -> Foo > >and can have "type variables", for instance: > >data Foo a = Bar | Baz a > >here, something of type "Foo a" is either of the form Bar or of the form >Baz x for some x which is of type a. This has constructors: > >Bar :: Foo a Baz :: a -> Foo a > >I hope this sheds some light on the issue... > > >-- >Hal Daume III | hdaume(a)isi.edu "Arrest this man, he talks in maths." | >www.isi.edu/~hdaume > _________________________________________________________________ Hot chart ringtones and polyphonics. Go to http://ninemsn.com.au/mobilemania/default.asp

1 0

Data types basics
by Patty Fong 05 Nov '03

05 Nov '03

2 1

Modeling multiple inheritance
by Brandon Michael Moore 05 Nov '03

05 Nov '03

I'm trying to build a nicer interface over the one generated by jvm-bridge. I'm using fancy type classes to remove the need to mangle method names. I would like methods to be automatcially inherited, following an inheritance hierarcy defined with another set of type classes. My basic classes look like this class HasFooMethod cls args result | cls args -> result where foo :: cls -> args -> result If I have classes A and B with foo methods like foo_JA_Jint :: ClassA -> Jint -> Bool foo_JB_Jboolean :: ClassB -> Bool -> Jint then I can make instances instance HasFooMethod ClassA Jint Bool instance HasFooMethod ClassB Bool Jint Now I can just use foo everywhere. I would like to avoid declaring an instance for every class though. In java methods are inherited from a superclass, and I would like to inherit methods automatically as well. In the bindings jvm-bridge generates a method is invoked with a function mangled after the highest ancestor that defined that particular overloading, so the implementation of HasFooMethod at a particular overloading is the same for any descendant. So I defined a class to model the inheritance relationships class SubType super sub | sub -> super where upCast :: sub -> super Now I can define a default instance of HasFooMethod: instance (HasFooMethod super args result, SubClass super sub) => HasFooMethod sub args result where foo sub args = foo (upCast sub) args This will propagate foo methods down the inheritance hierarcy. If a new class C is derived from A, I just need to say instance SubClass ClassA ClassC and ClassC gets a foo method. (In the actually code I piggy-back on a transitive subclass relation jvm-bridge defines that already includes an upcast method, so upCast has a default that should always be acceptable). The problem comes when interfaces are added to the mix. Interfaces are treated just like classes by jvm-bridge, and even though no implementation is inherited from instances in Java, the method accessors generated by jvm-bridge should be inherited. One problem is that the subclass relationship needs the functional dependency so that the default instance of HasFooMethod will respects the functional dependencies of HasFooMethod, so I can't declare subclass instances for multiple inheritance. On the other hand, if I don't use the functional dependency on HasFooMethod I end up needing to annotate most of the return values in a program. I run into similar problems trying to use numeric literals as arguments, because they are also overloaded. Does anyone know of clever solutions that would model multiple inheritance while preserving the functional dependencies (unsafe compiler flags are fine too), or ways to reduce the pain of overloading resolution without the functional dependency? One alternative is generating seperate HasFooMethod instances for every class in the system. The problem is that this would require alterating the bit of jvm-bridge that uses JNI to find information on classes, which currently only reports newly defined methods. JNI is black magic to me. Thanks Brandon

3 11

AW: AW: Heap profiling in GHC broken?
by Markus.Schnell＠infineon.com 04 Nov '03

04 Nov '03

> The canonical tutorial paper is in the Advanced Functional Programming > Summer Schools series: > > C. Runciman and N. Ro"jemo. Heap profiling for space efficiency. > In J. Launchbury, E. Meijer, and T. Sheard, editors, 2nd > Intl. School > on Advanced Functional Programming, pages 159-183, Olympia, WA, > August 1996. Springer LNCS Vol. 1129. Great! I knew only the more compiler-oriented Runciman/Röjemo papers. (Which were already helpful, but not very 'tutorial'.) Thanks, Markus

1 0

AW: Heap profiling in GHC broken?
by Markus.Schnell＠infineon.com 04 Nov '03

04 Nov '03

I have no idea how to set up cvs under windows to get things from a non-local repository, but anyway, profiling worked on the macintosh. I spent the whole weekend plus monday to find the space leak, but eventually I did. Heap Profiling was very helpful there. Unfortunately, space leaks are barely mentioned in the Haskell books/tutorials, but they are a real burden if you want to run programs for larger problems. A tutorial on this by one of the experts would be very welcome. :-) Thanks, Markus > -----Ursprüngliche Nachricht----- > Von: haskell-cafe-bounces(a)haskell.org > [mailto:haskell-cafe-bounces@haskell.org] Im Auftrag von David Roundy > Gesendet: Donnerstag, 30. Oktober 2003 13:48 > An: haskell-cafe(a)haskell.org > Betreff: Re: Heap profiling in GHC broken? > > > On Wed, Oct 29, 2003 at 05:43:23PM +0100, > Markus.Schnell(a)infineon.com wrote: > > When I tried to profile my program in search for space leaks, > > it core dumped on me (actually on Win2000/Cygwin). I'm > using ghc 6.0.1. > > Does anybody know what could be the problem? > > > > (To be more precise: any +RTS -h_ options make problems, > +RTS -p works > > fine. I get a message meaning the equivalent of "statement in > > '0x0058db43' points to memory in '0xfffffffc'. Could not read.") > > I believe this bug is fixed in CVS, so you could get it there > or wait for > 6.2. (I didn't fix it, I just reported it.) > -- > David Roundy > http://www.abridgegame.org/darcs > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe >

3 2