September 2010 - Generics

Fwd: [Haskell] ANNOUNCE: Utrecht Haskell Compiler (UHC), version 1.1.0
by José Pedro Magalhães 23 Sep '10

23 Sep '10

Hi generic programmers, I would like to emphasize the release of the Utrecht Haskell Compiler version 1.1.0---this time with built-in generics! Derivable type class functions are now implemented within Haskell, and you can have your own derivable type classes. In particular, you can also derive Functor, for instance. The mechanism is described in our paper soon to be presented at the Haskell Symposium [1]. So if you feel a bit experimental, try it out, and please report any problems you might run into. Cheers, Pedro [1] José Pedro Magalhães, Atze Dijkstra, Johan Jeuring, and Andres Löh. A generic deriving mechanism for Haskell. http://dreixel.net/research/pdf/gdmh_draft.pdf ---------- Forwarded message ---------- From: Atze Dijkstra <atze(a)cs.uu.nl> Date: Wed, Sep 22, 2010 at 16:30 Subject: [Haskell] ANNOUNCE: Utrecht Haskell Compiler (UHC), version 1.1.0 To: Haskell(a)haskell.org, haskell-cafe Cafe <haskell-cafe(a)haskell.org> Utrecht Haskell Compiler -- second release 1.1.0 ================================================ The UHC team is happy to announce the second public release of the Utrecht Haskell Compiler (UHC). UHC supports almost allmost all Haskell98 and Haskell2010 features plus some experimental extensions. The compiler runs on MacOSX, Windows (cygwin), and various Unix flavors. This second release has the following highlights relative to the previous first release: * Supports most of the Haskell2010 features. * Generic deriving [1]. * UHC knows about packages, and Cabal version 1.9.3 and later have basic support for UHC, so that Cabal can be used to build UHC packages. * For the default backend libraries are provided until including package haskell98. * The runtime system is no longer dependent on Boehm garbage collector, but uses UHCs own, a swap space collector allowing parameterization with backend specifics. * More programs from the nofib test suite compile and run. * Many smaller and larger problems fixed. For more info see http://www.cs.uu.nl/wiki/UHC Getting started & Download -------------------------- UHC is available for download as source distribution via the UHC home page: http://www.cs.uu.nl/wiki/UHC Here you will also find instructions to get started and the specifics of what UHC can and cannot do. Status of the implementation ---------------------------- Over the past year much effort by many people has been put into improving UHC, in particular to make it compile and run more library modules and test programs. Although the UHC project is very much work in progress overall reliability has improved, bugs have been fixed, and some Hackage packages compile. Warning ------- Although we think that the compiler is stable enough to compile substantial Haskell programs, we do not recommend yet to use it for any serious development work in Haskell. We ourselves use the GHC as a development platform! We think however that it provides a great platform for experimenting with language implementations, language extensions, etc. Mailing lists ------------- For UHC users and developers respectively: http://mail.cs.uu.nl/mailman/listinfo/uhc-users http://mail.cs.uu.nl/mailman/listinfo/uhc-developers Bug reporting ------------- Please report bugs at: http://code.google.com/p/uhc/issues/list The UHC Team [1] Jose Pedro Magalhaes, Atze Dijkstra, Johan Jeuring, and Andres Loeh. A generic deriving mechanism for Haskell. http://dreixel.net/research/pdf/gdmh_draft.pdf (see Section 7.1 for the discussion on constrained datatypes) - Atze - Atze Dijkstra, Department of Information and Computing Sciences. /|\ Utrecht University, PO Box 80089, 3508 TB Utrecht, Netherlands. / | \ Tel.: +31-30-2534118/1454 | WWW : http://www.cs.uu.nl/~atze<http://www.cs.uu.nl/%7Eatze>. /--| \ Fax : +31-30-2513971 .... | Email: atze(a)cs.uu.nl ............ / |___\ _______________________________________________ Haskell mailing list Haskell(a)haskell.org http://www.haskell.org/mailman/listinfo/haskell

1 0

make type witnesses 'phi ix' persistent?
by Sebastian Menge 20 Sep '10

20 Sep '10

Hi I am playing around with the zipper package. I wrote some helper functions that allow for a more convenient description of the zipper state as addresses in the basic form "path@term". The types and auxiliary functions are given below. Now I would like to make these addresses "persistent". I want to write them to disk somehow, and reconstruct them in a later session. But to reconstruct addresses, I have to create the type witnesses from the serialization, e.g.: serialize :: forall phi ix . (Fam phi, El phi ix) => phi ix -> String serialize MyWitness1 = "MyWitness1" serialize MyWitness2 = "MyWitness2" But deserialize wont work: deserialize :: String -> phi ix -- <--- impossible deserialize "MyWitness1" = MyWitness1 deserialize "MyWitness2" = MyWitness2 Is there _any_ way to get this going ? Any dirty hacks are also appreciated ;) Thanks, Sebastian. PS here come the relevant types and functions (not tested, just copy'n'pasted from my sources) -- more convenient zipper states data Address phi s t = Address { typerep :: phi s , value :: s , pathrep :: phi t , path :: Path } type Path = [Int] type Nav phi ix= (Loc phi I0 ix -> Maybe (Loc phi I0 ix)) -- convert int-paths to navigations tonav :: Path -> Nav phi ix tonav (path) = foldl (>=>) return (map tonav' path) where tonav' 0 = down tonav' i = tonav' (i-1) >=> right -- lookup an address lookup::(Fam phi, Zipper phi (PF phi), EqS phi) => Address phi s t -> t lookup a = (pathrep a, path a) @@ (typerep a, value a) -- explicit lookup on pairs. idea: "path@term" (@@) :: forall phi ix1 ix2. (Fam phi, Zipper phi (PF phi), EqS phi) => (phi ix2,Path) -> (phi ix1,ix1) -> ix2 (@@) (q,path) (p,e) = case (tonav path $ enter p e) of (Just l :: (Maybe (Loc phi I0 ix1))) -> focus q l Nothing -> error "invalid path" -- helper that extracts the focus focus :: forall phi ix xi. (EqS phi) => phi ix -> (Loc phi I0 xi) -> ix focus p (Loc q (I0 x) cs) = case eqS p q of Just Refl -> x _ -> error "focus: types dont match"

2 3

type constructor genericity with regular
by Jan Christiansen 15 Sep '10

15 Sep '10

Hi, I want to define a function which is generic for all unary type constructors using the regular package. More precisely, I want to define a function like the following, which replaces every polymorphic component by the position of the component in the generic representation. shape :: (Regular f, Shape (PF f)) => f a -> f [Bool] Is it correct that it is not possible to implement a function like this using regular? Or do I miss something? I have implemented the function by using different representation types K, I and so forth that take an additional argument. Furthermore I have added an additional type to mark the polymorphic component. Thanks, Jan

2 5

ANNOUNCE: grammar-combinators 0.1 (initial release): A parsing library of context-free grammar combinators
by Dominique Devriese 09 Sep '10

09 Sep '10

Copying this announcement to generics(a)haskell.org. The grammar-combinators library may interest some people here, because it uses Multirec pattern functors throughout, to abstract a grammar from its semantics. More details in the paper we are submitting to PADL 2011 [1]. Dominique [1]: http://projects.haskell.org/grammar-combinators/#background The grammar-combinators library is a parsing library employing a novel grammar representation with explicit recursion. The library features much of the power of a parser generator like Happy or ANTLR, but with the library approach and most of the benefits of a parser combinator library. Grammars and grammar algorithms are defined in a functional style. The library currently has the following features: * Grammar specified completely in Haskell using an elegant syntax * Grammar algorithms implemented in a functional style (no fresh identifiers), with elegant and meaningful types. * Multi-backend: use the same grammar with a Packrat, Parsec or UUParse parser * Grammar transformations: use left-recursive grammars directly thanks to a powerful grammar transformation library, featuring the left-corner left-recursion removal transform, a uniform version of the classic Paull left-recursion removal, and various smaller transformations (dead-branch removal, dead non-terminal removal, consecutive epsilon combination, selective unfolding etc.). * Grammar utility functions: printing of grammars, FIRST-set calculation, reachability analysis of non-terminals, etc. * Compile-time transformations (using Template Haskell), given a suitable definition of the grammar. This is currently limited to a certain set of transformations. The library is currently not intended for mainstream use. Its API is relatively stable, but performance needs to be looked at further. We are submitting a paper about the ideas behind this library to PADL 2011. A draft is linked on the project's website. More information: * Project website: http://projects.haskell.org/grammar-combinators/ * Tutorial: http://projects.haskell.org/grammar-combinators/tutorial.html * Hackage: http://hackage.haskell.org/package/grammar-combinators All comments welcome! Dominique PS. The documentation on hackage currently doesn't build because of (seemingly) a Hackage dependency problem during the build [1]. Compiling and generating the documentation locally should work fine. A version of the docs is available on the project's webpage as a temporary replacement [2]. Footnotes: [1] http://www.haskell.org/pipermail/libraries/2010-September/014168.html [2] http://projects.haskell.org/grammar-combinators/docs/index.html

1 0

darcs patch: update package description with new upper limit on the...
by Antoine Latter 07 Sep '10

07 Sep '10

1 patch for repository http://code.haskell.org/syb: Mon Sep 6 12:56:43 CDT 2010 Antoine Latter <aslatter(a)gmail.com> * update package description with new upper limit on the base dependency With this change I seem to be able to compile with GHC HEAD. The new Cabal issue wanrings directed us to use the new testing infrastructure, but that warning is left unadressed.

2 2

SYB with base >= 4.3
by Kathleen Fisher 03 Sep '10

03 Sep '10

Hi, I'd like to install the SYB package, but my base library is 4.3.0.0 (the version in the current GHC HEAD). The constraints on syb installation require base version < 4.3. Is there a new version somewhere that compiles against 4.3 that I could install? Kathleen

3 4