June 2006 - Haskell - Haskell.org

Forcing Type Class Equality
by john lask 29 Jun '06

29 Jun '06

** this is literate haskell hoping someone can help me. What I am trying to do is >class Ba1 a >ba1 :: Ba1 a => a -> IO () >ba1 x = print "helllo" what I wish to do is declare another function >class Foo a >proxy :: Foo a => a -> IO () so that I can do something like >proxy x = ba1 x this does not work in hugs and gives ERROR "test.lhs":19 - Cannot justify constraints in explicitly typed binding *** Expression : proxy *** Type : Foo a => a -> IO () *** Given context : Foo a *** Constraints : Ba1 a which is entirely expected. The trick is to tell hugs that Foo and Ba1 are the same so how do I do that ? I have tried >instance Ba1 a => Foo a >instance Foo a => Ba1 a which should define equality but to no avail ?? Can this be done ? I thought the problem might be drawing the inference Foo == Ba1 and to try and force the inference i did: >class Foo a >proxy1 :: (Foo a,Ba1 a) => a -> IO () >proxy1 x = ba1 x >proxy2 :: Foo a => a -> IO () >proxy2 x = proxy x This typed checked but Main> :load test Main> proxy2 ERROR - *** The type checker has reached the cutoff limit while trying to *** determine whether: *** Foo a *** can be deduced from: *** () *** This may indicate that the problem is undecidable. However, *** you may still try to increase the cutoff limit using the -c *** option and then try again. (The current setting is -c40) any ideas out there ?? _________________________________________________________________ Meet Sexy Singles today @ Lavalife - Click here http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Flavalife9%2Eninemsn%2Ecom%2…

2 2

ANN: MissingH 0.14.4
by John Goerzen 28 Jun '06

28 Jun '06

I am pleased to announce a new version of MissingH. Changes since the last announcement include: * Fix a bug in ConfigParser.merge and write a test for it. * Fix cabal incompatibilies with GHC 6.4.2 * IO/StatCompat: Fixed missing Data.Bits import on mingw32 * Updated COPYING file with new FSF address * Cmd.hs: Commented out code that was never used * Import new package MD5 0.2.7 from Ian Lynagh * ConfigParser gives more helpful error messages when possible -- it now includes section and option in most error messages * Unit tests updated for new ConfigParser output * MissingH.Str: New function escapeRe that takes a string and makes a regular expression that matches the string literally. * MissingH.IO.HVFS: New function vDoesExist * MissingH.List: New function hasAny * New modules: MissingH.Path.Glob and MissingH.Path.WildMatch and tests for them * 716 total unit tests now MissingH is available for download from http://quux.org/devel/missingh It is also included with several Linux and BSD distributions.

1 0

Automatic installation of cabalized librraies
by Bulat Ziganshin 28 Jun '06

28 Jun '06

Hello Jun, Wednesday, June 28, 2006, 9:58:11 AM, you wrote: > These bindings are already cabalized, you have to type only as follows > to install them: > % runghc ./Setup.hs configure > % runghc ./Setup.hs build > # runghc ./Setup.hs install i propose to use the Makefile attached in all cabalized libs. it allows to install library by just "make install" (if you don't received attached file, you can borrow it from http://www.haskell.org/library/Streams.tar.gz ) -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

1 0

ANNOUNCE: hmp3 1.1
by dons＠cse.unsw.edu.au 28 Jun '06

28 Jun '06

A new version of hmp3 has been released, version 1.1 hmp3 is a curses-based mp3 player written in Haskell. It is designed to be simple, fast and robust. This is mostly a maintenance, release, fixing support for ghc 6.4.2. However, you do get some new features: * Enable searching of the entire track list, not just the directory list * Adds ability to play previous song (thanks to Christophe Poucet) * Ported to use the Data.ByteString api Available as a cabalised src tarball, a darcs repo, and binaries for (currently): * Linux/x86 * OpenBSD/x86 It is also known to run on FreeBSD, OSX and Irix/mips64 and probably any other unix system with a minimal curses implementation. http://www.cse.unsw.edu.au/~dons/hmp3.html Cheers, Don

1 0

Graphing community activity
by dons＠cse.unsw.edu.au 28 Jun '06

28 Jun '06

Hey all, Inspired by a comment from Shae Erisson, I added a little 5 line script to darcs-graph[1], to display the commit activity of _remote_ darcs repositories. Here are the activity graphs for a selection of projects in the community: http://www.cse.unsw.edu.au/~dons/images/commits/community/ The graphs show, for instance, ghc activity back to 1996. This might be useful to quickly work out what's going on in the community. If you're interested in hacking something up, using this data, ping me :) A page showing hot projects, for example... -- Don 1. http://www.cse.unsw.edu.au/~dons/darcs-graph.html

1 1

Google SoC: Software Transactional Memory for Parrot
by Bulat Ziganshin 27 Jun '06

27 Jun '06

Hello haskell, browsing Google SoC site, i found one accepted project that you may be interesting to hear about: Software Transactional Memory for Parrot by Charles Albert Reiss, mentored by Leopold Toetsch (mentioned on http://code.google.com/soc/tpf/about.html ) it seems that Haskell continues to be a source of new technologies for other languages. i will wait for GADT for C# :) -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

6 5

haxr troubles - solved
by Johannes Waldmann 27 Jun '06

27 Jun '06

Björn has just made a new release http://www.haskell.org/haxr/download/haxr-20060626.tar.gz that solves the problems I mentioned. Thanks! -- -- Johannes Waldmann -- Tel/Fax (0341) 3076 6479/80 -- ---- http://www.imn.htwk-leipzig.de/~waldmann/ -------

1 0

ANNOUNCE: HSP.Clientside 0.001
by Joel Björnson 26 Jun '06

26 Jun '06

Hi. I would like to announce the *prerelease* of Hsp.Clientside. This is a 'Summer of Code' project aiming to add support for client-side script generation in Haskell Server Pages. ( http://www.cs.chalmers.se/~d00nibro/hsp/) At the moment the basic building blocks for embedding Javasript has been implemented. As the project proceeds a suitable programming model based on these components will be added. Hopefully this will also include some kind of higher level Ajax support. If you are interested please take a look at darcs.haskell.org/SoC/hsp.clientside. Off course I appreciate all sorts of comments, opinions and proposals regarding this work. For more information see www.dtek.chalmers.se/~bjornson/soc /Joel

1 0

haxr (XML RPC server) problems with C# client
by Johannes Waldmann 26 Jun '06

26 Jun '06

Hello. I am trying to access an haxr (XML RPC) server from a C# client, see http://www.imn.htwk-leipzig.de/~waldmann/edu/ss06/oo/folien/oo/node153.html and it does not work: on the client side, I get Unhandled Exception: CookComputing.XmlRpc.XmlRpcFaultException: Server returned a fault exception. in <0x00274> CookComputing.XmlRpc.XmlRpcSerializer:DeserializeResponse (System.Xml.XmlDocument xdoc,System.Type returnType) How do I find out more (e. g. the server's actual response)? The same server works with a Java client. Is there a way to have a standalone haxr server, i. e. a process listening on a port, without the detour along the web server/cgi mechanism? (cf. the Java variant) Thanks, -- -- Johannes Waldmann -- Tel/Fax (0341) 3076 6479/80 -- ---- http://www.imn.htwk-leipzig.de/~waldmann/ -------

1 1

extensible records using associated types
by Barney Hilken 25 Jun '06

25 Jun '06

I'm sorry if this example has already been posted, but I couldn't find it by searching the archives. If it's new (and works properly), it provides more evidence that ATs are a good thing. You can use associated type synonyms to implement polymorphic extensible records. ------------------------------------------------------------------------ --------- We will represent records as values of the form N1 x1 (N2 x2 ... (Nn xn Empty)...) with corresponding type N1 a1 (N2 a2 ... (Nn an Empty)...) where the field name N is defined by the declaration data N a r = N a r For convenience, we define the type of field name constructors: > type Constructor n = forall a. forall r. a -> r -> n a r so that N :: Constructor N for each field name N. Next we define a class (:<:) whose purpose is to tell the compiler in which order to store fields: > class (m :: * -> * -> *) :<: (n :: * -> * -> *) The idea is that if N and M are distinct field names which might appear the same record, then there must be an instance of exactly one of N :<: M and M :<: N. The class Contains n r asserts that r is a record type with a field labelled n: > class Contains n r where > type Project n r > type Delete n r > project :: Constructor n -> r -> Project n r > delete :: Constructor n -> r -> Delete n r There are two associated types: Project n r is the type of the n field, and Delete n r is the type of the rest of the record. The functions project and delete give the corresponding values. The class Lacks n r asserts that r is a record type without a field labelled n: > class Lacks n r where > type Extend n a r > extend :: Constructor n -> a -> r -> Extend n a r There is one associated type: Extend n a r is the type of the record obtained by adding a field of type a labelled n. Note that according to the AT paper, the parameter a should come last, but this order is much more convenient. The function extend adds the new field to the record. The class Disjoint r s asserts that r and s are record types with no fields in common: > class Disjoint r s where > type Union r s > union :: r -> s -> Union r s There is one associated type: Union r s is the type of the merged record, and union is the merge function. Next we define the empty record: > data Empty = Empty It has no fields, so has no instances of Contains, but lots of instances of Lacks: > instance Lacks n Empty where > type Extend n a Empty = n a Empty > extend nn x Empty = nn x Empty It is also disjoint from everything: > instance Disjoint Empty r where > type Union Empty r = r > union Empty t = t Each field name must be defined as a data type together with instances of the classes Contains, Lacks and Disjoint as follows: > data N a r = N a r The type N a r certainly contains a field named N: > instance Contains N (N a r) where > type Project N (N a r) = a > type Delete N (N a r) = r > project N (N x t) = x > delete N (N x t) = t It also Contains all the fields of r, though you only need to check names which are below N: > instance m :<: N, Contains m r => Contains m (N a r) where > type Project m (N a r) = Project m r > type Delete m (N a r) = N a (Delete m r) > project mm (N x t) = project mm t > delete mm (N x t) = N x (delete mm t) Similarly, it Lacks all the fields which r Lacks, except for N itself. This is where we really need the ordering :<: to ensure that m is not equal to N. There are two cases, m :<: N: > instance m :<: N, Lacks m r => Lacks m (N a r) where > type Extend m b (N a r) = N a (Extend m b r) > extend mm y (N x t) = N x (extend mm y t) and N :<: m: > instance N :<: m => Lacks m (N a r) where > type Extend m b (N a r) = m b (N a r) > extend mm y (N x t) = mm y (N x t) To ensure that N a r is disjoint from s, it is enough that r is disjoint from s and N is not in their union: > instance Disjoint r s, Lacks N (Union r s) => Disjoint (N a r) s where > type Union (N a r) s = Extend N a (Union r s) > union (N x t) u = extend N x (union t u) Finally, we must maintain the ordering :<:. It has to be a global linear order, so if we know the last field name was M, we can define: > instance M :<: N > instance m :<: M => m :<: N Usage. ------ The record {N1 = x1, ... Nn = xn} should be constructed as (extend N1 x1 $ ... $ extend Nn xn) Empty. Extension, projection, deletion and union are all as polymorphic as you could want, and you can even use field names as ordinary values. For example, you can define update by: > update :: Contains n r, Lacks n (Delete n r) => Constructor n -> a -> r -> Extend n a (Delete n r) > update n y t = extend n y (delete n t) Problems. --------- 1. Pattern matching. Pattern matching on complete records is possible (though you must get the fields in the right order!) but there is no equivalent to the trex .. notation for 'any other fields'. 2. Maintaining the :<: order. Within one module this could be generated with a template, but between modules this is more of a problem. Of course, we could use a built-in class based on alphabetical order, but this would need compiler support. 3. Controlling the use of constructors. If you use N to construct records and types directly, instead of Extend N and extend N, you get fields in the wrong order and everything breaks. The type system can't stop you doing this. 4. Testing. I don't actually know if this works, because phrac won't compile it. We really need the second-order polymorphism of field constructors, and phrac doesn't have forall. (There is also the trivial point of the order of arguments to Extend). Barney.

2 4