March 2007 - Haskell-Cafe

Newbie question: ghc search path for .c files
by Seth J. Fogarty 16 Mar '07

16 Mar '07

Hi, I am having difficulty compiling a program that uses a foreign library. The readme says to first compile a C file: :> ghc mipFFI.c -c mipFFI.c:1:0: lpkit.h: No such file or directory Alright, that's not uncommon. gcc confirms :> gcc mipFFI.c mipFFI.c:1: lpkit.h: No such file or directory mipFFI.c:2: HsFFI.h: No such file or directory So I then include the directory and gcc finds the lpkit.h file. :> gcc -I /home/sfogarty/lib/lp_solve_5.5 mipFFI.c mipFFI.c:2: HsFFI.h: No such file or directory However ghc does not seem to accept this :> ghc -i/home/sfogarty/lib/lp_solve_5.5 mipFFI.c mipFFI.c:1:0: lpkit.h: No such file or directory I've tried every variation of spacing, capitalization, copying the files to the local directory, simply listing the files ahead. Nothing seems to work! I am horribly confused! And another question while I am here: -package posix and -package lang don't seem to be included on my distribution of ghc. I am having minimal success, as yet, finding where to download them. If I could get a pointer, I'd be very appreciative. If not, I'm sure I'll find it eventually. -- Seth Fogarty sfogarty(a)[gmail.com|rice.edu|livejournal] Neep-neep at large AIM: Sorrath "Let us not seek to satisfy our thirst for freedom by drinking from the cup of bitterness and hatred." - Martin Luther King, Jr.

2 1

Lack of expressiveness in kinds?
by Andrew Wagner 16 Mar '07

16 Mar '07

Ok, so I'm inching closer to understanding monads, and this question popped up today. Consider the following 2 declarations: data Foo a = Bar a data (Ord a) => Baz a = Bah a Note that both of these have kind * -> *. However, Baz could never be an instance of monad, because there is a restriction on the types it can operate on. Foo, however, is completely polymorphic, without limitation. It seems to me that there ought to be a way to express the difference between the two in the type/kind system. For example, you can almost, but not quite, say that in the declaration "class Monad m where..", m must be of kind *->*, but that's not quite enough to say, because of this example. Am I just missing something, or is there a reason the kind of Baz shouldn't be something other than *->*?

3 2

ANNOUNCE: Google Summer of Code - open for business
by Malcolm Wallace 16 Mar '07

16 Mar '07

We are pleased to announce that haskell.org has been officially accepted as a mentoring organisation for this year's Summer of Code. Student applications are now open! http://code.google.com/summerofcode.html http://hackage.haskell.org/trac/summer-of-code/ Students: want to earn $4500 for 3 months work hacking in Haskell? Go to our ideas page, browse around, and write up a proposal for something you would like to work on. You must be available to work basically full-time for the months of June to August (inclusive). Proposals are competitive against other students. Last year, nine were funded. Your deadline for application is 24th March. Anyone else: Got any good ideas for coding projects, but no time to work on them yourself? Go to our wiki and add it in to the database! Or do you feel you could be either a primary or backup mentor for a student project in an area of your interest, even if it isn't a project you defined yourself? We need willing members of the community to foster the new student talent, so sign up on the wiki's 'people' page if you would like to be involved. Obviously you will need Haskell coding experience, but also be able to advise on time management, strategies for gaining a userbase and keeping them happy, ways of interacting with the wider Haskell community, and so on. Regards, Malcolm

1 0

Typical?
by Hans van Thiel 16 Mar '07

16 Mar '07

Hello All, Us who're learning Haskell have a lot in common... that is, if Amazon has it right... ------------ We recommend: STAR TREK-USS ENTERPRISE D by Corgi Classics Recommended because you purchased or rated: * The Haskell School of Expression: Learning Functional Programming through Multimedia . . . We recommend: Cooking Entrees With the Micheff Sisters: A Vegan Vegetarian Cookbook by Brenda Micheff Walsh Recommended because you purchased or rated: * Purely Functional Data Structures ------------ To be fair, other recommendations were more traditional ...:) Hans van Thiel

2 1

Generalised merge
by Paul Johnson 16 Mar '07

16 Mar '07

The most common kind of primitive recursive function I find myself writing these days is a variation on the theme of merging two sorted lists. You can see some examples in my Ranged Sets library at http://ranged-sets.sourceforge.net/. For instance: -- | Set union for ranged sets. Infix precedence is left 6. rSetUnion, (-\/-) :: DiscreteOrdered v => RSet v -> RSet v -> RSet v -- Implementation note: rSetUnion merges the two lists into a single -- sorted list and then calls normalise to combine overlapping ranges. rSetUnion (RSet ls1) (RSet ls2) = RSet $ normalise $ merge ls1 ls2 where merge ls1 [] = ls1 merge [] ls2 = ls2 merge ls1@(h1:t1) ls2@(h2:t2) = if h1 < h2 then h1 : merge t1 ls2 else h2 : merge ls1 t2 -- | Set intersection for ranged sets. Infix precedence is left 7. rSetIntersection, (-/\-) :: DiscreteOrdered v => RSet v -> RSet v -> RSet v rSetIntersection (RSet ls1) (RSet ls2) = RSet $ filter (not . rangeIsEmpty) $ merge ls1 ls2 where merge ls1@(h1:t1) ls2@(h2:t2) = rangeIntersection h1 h2 : if rangeUpper h1 < rangeUpper h2 then merge t1 ls2 else merge ls1 t2 merge _ _ = [] Union also has its own merge function. The worst case I've come across was at work. I can't talk about the details, but it involved manipulating two functions of time represented by lists of samples. So I had a type TimeFunc = [(Value, Time)], and the job was to compare two TimeFuncs with samples at different times. Step 1 was to interpolate each TimeFunc with the values for the times in the other TimeFunc, giving a result of type [(Value, Value, Time)] for the union of all the times in both original TimeFuncs. I wrote a truly hairy zipTimeFunc function with guards to match each possible case. It worked, but it must have been 100 lines if you include the comments to explain each case and demonstrate totality. So I'm wondering if anyone has a more general pattern. Unfold? Some variation on a theme of zip? I once tried writing a generalised merge, but it needed half a dozen functions as arguments to handle all the various cases. It was kludgier than just rolling a new merge routine every time. And I don't think it would have handled the TimeFunc case. Paul.

2 1

Haskell beginner
by arnuld 15 Mar '07

15 Mar '07

i want to start learning Haskell and willing to master it :-) i have done some Common Lisp from http://www.gigamonkeys.com/book/. so i know what are functions, variables, etc. BUT i have never done any real life programmming or any kind of software development. in my country no Haskell books are available :-( , hence i can only use online Tutorials. my main purpose is to learn real-life programming by contributing some "Haskell coding" to a GPLed Software written using Haskell. i have found "Yet Another Haskell Tutorial" after searching the archives. is it really a good idea to learn Haskell using only Online material ? (when one does not have any offline resources) -- http://arnuld.blogspot.com/

8 7

How do I avoid stack overflows?
by DavidA 15 Mar '07

15 Mar '07

Hi. I'm trying to write some code which involves lots of matrix multiplications, but whenever I do too many, I get stack overflows (in both GHCi 6.4.2, and Hugs May 2006). The following code exhibits the problem. import List (transpose) u <.> v = sum $ zipWith (*) u v a <<*>> b = multMx a (transpose b) where multMx [] _ = [] multMx (u:us) bT = map (u <.>) bT : multMx us bT id3 = [[1,0,0],[0,1,0],[0,0,1]] test = iterate (<<*>> id3) id3 !! 1000000 I tried to fix the problem using seq, as follows: iterate' f x = x : seq x' (iterate' f x') where x' = f x test' = iterate' (<<*>> id3) id3 !! 1000000 However, in both cases, the code causes stack overflows in both interpreters. (It sometimes kills GHCi, which I guess is a bug.) Any ideas? Thanks.

2 1

Foralls in records
by Adde 15 Mar '07

15 Mar '07

Hi. I'm experimenting with implementing database transactions as monads but I'm getting stuck on how to store a generic connection (only constrained by a typeclass) inside the transaction. The reason I'm doing it this way is that the connection could be a different kind of structure depending on what database the transaction is using. The goal is to enable all functions using the monad to extract the connection using getConnection. I realise I could probably get it running by adding a type parameter for the connection type to TransactionT and Transaction but I feel I shouldn't have to. All the compiler needs to know is that the contained value conforms to the interface of Connection. <code> data TransactionT = forall c. (Connection c) => TransactionT c data Transaction a = Transaction (TransactionT -> (a, TransactionT)) instance Monad Transaction where ... getConnection :: Transaction c getConnection = Transaction (\t@(TransactionT c) -> (c, t)) class Connection c where connectionExecute :: c -> String -> Transaction () execute :: String -> Transaction () execute s = connectionExecute getConnection s </code> I'm getting the following error: Couldn't match expected type `c' (a rigid variable) against inferred type `c1' (a rigid variable) `c' is bound by the type signature for `getConnection' at Database.hs:19:29 `c1' is bound by the pattern for `TransactionT' at Database.hs:20:33-46 In the expression: c In the expression: (c, t) In a lambda abstraction: \ (t@(TransactionT c)) -> (c, t)

5 11

Make strict (IO String) lazy
by Henning Thielemann 15 Mar '07

15 Mar '07

On the one hand, in the standard libraries there are functions like readFile, getContents, hGetContents which read a file lazily. This is often a nice feature, but sometimes lead to unexpected results, say when reading a file and overwriting it with modified contents. Unfortunately the standard libraries provide no functions for strict reading, and one has to do this manually. On the other hand, when I write some IO function that returns a String, I easily end up with a function which produces the String in a strict way. (Say I call some shell commands and concatenate their outputs.) What is the preferred way to turn a strict (IO String) into a lazy one? forkIO? forkOS? How would one derive readFile from a hypothetical strictReadFile?

3 2

SYB vs HList (again)
by S. Alexander Jacobson 15 Mar '07

15 Mar '07

Right now I am looking at using either SYB (Scrap Your Boilerplate) or HList Records to eliminate boilerplate for: * parsing URLEncoded strings into application data structures * generating XML/JSON from application data structures * handling adding new fields to serialized data structures * creating indexed collections of application data structures I am looking for insights from people here on which approach they think is better and why. Here are my current thoughts on the issue == Both HList and SYB require data structures redesign == HLists require you to define Labels and basically only use label values that are themselves either scalar or HLists. SYB basically requires the same thing except that you use data/newtype to define labels instead of HLists more cumbersome label constructions. == Defaults: HList gives you compile time errors, SYB only runtime errors == SYB does not seem to provide a way of having the compiler tell you that you are accessing a field that is unavailable in the type. HList will give you a type error if you do that. I don't know exactly how HList handles default values but I assume you can restrict use of those values to explicit deserialization contexts. Is that correct? == HList allows more informality == With HList you can specify the type of a record ad hoc using obj::(Record Foo .*. Bar .*.Baz). SYB requires that you define data structures in separate data declarations. It would be really nice if there was some way to tell Haskell that HLists have no more fields than the ones you happen to be getting and setting in your code. Effectively that would mean you get data structure inference not just function type inference which would be really cool! That is probably not possible but it couldn't hurt to ask (Oleg?). == SYB doesn't require template haskell to make it usable == With SYB you create field labels using newtype (or data) declarations e.g. data Salary = S {salary::Float} With HList, label declarations are really verbose e.g. data SalaryLabel deriving(Typeable) type Salary = Field (Proxy SalaryLabel) Int salary = proxy :: Proxy FooLabel You can make this more concise using TemplateHaskell but TH looks alien and adds fear to the use of any code. == Performance issues == SYB requires a linear traversal of all field elements using dynamic to get or transform a value. HList traverses an HCons list. I don't how bad this is as compared with traditional data structure access using pattern matching or field labels. My current bias is towards using HList because if we are going to force a conversion to a new data structure convention I'd rather have the typesystem on my side. With SYB it is too easy to let haskell field labels creep in to your data definitions and end up with subtle errors to correct. Any opinions on these issues would be very appreciated. -Alex-

2 1