- Haskell - Haskell.org

Data.Map missing keysFM_GE, eltsFM_GE?
by S. Alexander Jacobson 21 Feb '05

21 Feb '05

Data.FiniteMap has functions keysFM_GE and eltsFM_GE which return the keys or elts of pairs greater than or equal to the provided key. These functions do not appear to be in the new Data.Map. Is this intentional? -Alex- ______________________________________________________________ S. Alexander Jacobson tel:917-770-6565 http://alexjacobson.com

2 2

Call for IPSI Transactions Special Issues in 2005/6; c/ba
by IPSI Transactions Special Issues 21 Feb '05

21 Feb '05

Dear potential Speaker: We are pleased to inform you that both IPSI Transactions journals are planing some special issues in late 2005 and early 2006, and you are welcome to submit your paper(s), until the deadlines listed below! IPSI Transactions on Internet Research: March 31, 2005 - Special Issue on E-Education: Concepts and Infrastructure June 30, 2005 - Special Issue on E-Business: Concepts and Infrastructure IPSI Transactions on Advanced Research: March 31, 2005 - Special Issue on the Research with Multidisciplinary Elements June 30, 2005 - Special Issue on the Research with Interdisciplinary Elements Each submitted paper first undergoes the editor review, and those who pass this first stage are sent to 12 external experts for a rigorous review; decisions are made after at least 6 external reviewers respond! The review is free of charge, but the authors of the accepted papers ar expected to pay the publication fee of E400 per paper (if 4 or 5 or 6 pages of the TIR/TAR format), and the additional fee of E100 per page, for each extra page, till the maximum of 10 pages. Rigorous reviewing is the major strength of IPSI journals, which is the major contributor to their high quality! Soft copies of the existing issues of TIR and TAR can be seen at the web, and hard copies can be obtained on a special request by email, as indicated on the web, where can find all information! Sincerely yours, Prof. Dr. Veljko Milutinovic, Editor-in-Chief P.S. If you need aditional information, please reply to this e-mail.

1 0

ANN: Hacanon 0.1
by Lemmih 21 Feb '05

21 Feb '05

What is Hacanon? Hacanon is a Template Haskell library which automates the process of binding Haskell to libraries written in C++. It's designed to capture common patterns and allow the programmer to specify the desired behaviour accordingly. Hacanon is capable of generating a usable binding, without any interaction, to almost any piece of C++. This binding is of relatively high level, because Hacanon translates enumerations to Haskell data structures, has transparent callback support and overloads every function. Unfortunately, such a binding would, in most cases, be flawed. Is 'int*' an array, variable which holds a second result or something third? Here Hacanon provides a data interface scheme (aka DIS) system which allows the programmer to concisely describe how an argument should be handled. DISs can be derived from other DISs because they're nothing more than data structures. This gives rise to increased code reuse and minimizes the burden on the programmer. These DISs can be used in the appliance system where the programmer can guard on function name (via regex), argument type and argument number. Some of the features sound familiar. Is Hacanon related to GreenCard or HDirect? I've borrowed the DIS concept from GreenCard and directional arguments from HDirect. I would like to think I've taken the best of both worlds and added some innovation in form of automation. Dependencies: * Recent version of GHC. Tested with ghc-6.4.20050209 and ghc-6.4.20050219. http://www.haskell.org/ghc * HaXml 1.12. http://www.cs.york.ac.uk/fp/HaXml/ * GCCXML from CVS hacked to include template information. Hack is distributed with Hacanon. Author and availability: Hacanon is written by David Himmelstrup and published under a MIT-style license. Darcs repository: http://www.scannedinavian.org/repos/hacanon/ -- Friendly, Lemmih

2 2

Midlands Graduate School
by N Ghani 21 Feb '05

21 Feb '05

Dear Functional Programmers On 11-15 April, at the University of Birmingham, the Midlands Graduate School (a consortium of theoretical CS researchers at Birmingham, Nottingham, Leicester and Warwick) will be holding a spring school. It is aimed at first-year PhD students, but is open to other participants as well. http://www.cs.bham.ac.uk/~pbl/mgs/ The core courses will be Category theory, Graham Hutton Typed lambda-calculus, Paul Levy Denotational semantics, Achim Jung Students can select any combination of the following optional courses that fits the schedule. Programming stream Quantum programming, Thorsten Altenkirch Epigram, Conor McBride Functional reactive programming, Henrik Nilsson Logic and model-checking stream Modal logic, Natasha Alechina Game semantics for software verification, Dan Ghica Verifying security protocols, Mark Ryan Semantics stream Operational domain theory and topology, Martin Escardo Abstract Stone duality, Paul Taylor Game semantics for software verification, Dan Ghica And all that for just £40. University accommodation (£25 per night) can also be booked. Too much? Funding is available from EPSRC (UK) and APPSEM (EU), see http://www.cs.bham.ac.uk/~pbl/mgs/ for more details, and for a registration form. See you there Paul P.S. Students will also be able to attend a Workshop on Logic and Contro l on 16 April, for an additional fee. Details yet to be finalized.

1 0

[announce] Fedora Haskell mailing list
by Jens Petersen 21 Feb '05

21 Feb '05

The Fedora Haskell project now has a mailing list which is kindly hosted on haskell.org. :) If you wish to subscribe for announcements and discussion of packaging Haskell projects for Fedora, please go to <http://haskell.org/mailman/listinfo/fedora-haskell>. More information about the project is available on <http://haskell.org/fedora> Cheers, Jens

1 0

ICLP 2005: Call for Papers
by ICLP 2005 21 Feb '05

21 Feb '05

---------------------------------------- Apologies if you receive multiple copies ________________________________________ Call for Papers ------------------------------------------------------------ Twenty first International Conference on Logic Programming ICLP'05 2-5 October, 2005 Sitges (Barcelona) Spain Co-located with the International Conference on Principles and Practice of Constraint Programming (CP'05) ------------------------------------------------------------ http://www.iiia.csic.es/iclp2005/ The Conference -------------- The 21st International Conference on Logic Programming will be held near Barcelona (Spain) from October 2nd to October 5th, 2005. ICLP'05 will be colocated with the International Conference on Principles and Practice of Constraint Programming (CP'05). Conference scope ---------------- Since the first conference held in Marseilles in 1982, ICLP has been the premier international conference for presenting research in logic programming. Contributions (papers and posters) are sought in all areas of logic programming including but not restricted to: Theory Implementation Semantic Foundations Compilation Formalisms Memory Management Nonmonotonic Reasoning Virtual Machines Knowledge Representation Parallelism Environments Alternative Paradigms Program Analysis Constraint Logic Programming Program Transformation Abductive Logic Programming Validation and Verification Inductive Logic Programming Debugging, Profiling Answer Set Programming Language Issues Applications Concurrency Semantic Web Objects Software Engineering Coordination Web Tools Mobility Internet Agents Higher Order Artificial Intelligence Types Deductive Databases Modes Natural Language Programming Techniques Specific attention will be given to work providing novel integrations of these different areas, and to new applications of logic programming in general. Contributions on applications will be assessed with an emphasis on their impact and synergy with other areas, as opposed to technical maturity. Applications of logic programming to the Semantic Web are especially encouraged. The technical program will include several invited talks and advanced tutorials, in addition to the presentations of the accepted papers and posters. A special session on industrial applications of logic programming is also planned and several workshops will be held in parallel with the conference. For the first time, a doctoral student consortium will be organized as part of ICLP. Papers ------- Papers must describe original, previously unpublished research, and must not be simultaneously submitted for publication elsewhere. They must be written in English and not exceed 15 pages in Springer LNCS format. The authors are encouraged, although not obliged, to submit their papers already in Springer LNCS format. General information about the Springer LNCS series and the LNCS authors' instructions are available at the Springer LNCS/LNAI home page (http://www.springer.de/comp/lncs/index.html) Papers should express their contribution clearly, both in general and technical terms. It is essential to identify what was accomplished, describe its significance, and explain how the paper compares with and advances previous work. Authors should make every effort to make the technical content understandable to a broad audience. The primary means of submission will be electronic, in pdf format. If electronic submission is not possible, five hard copies should be sent to one of the program co-chairs. More information on the submission procedure will be available at http://www.utdallas.edu/ICLP05 Industrial Papers ----------------- A special session on industrial applications of logic programming is also planned during the conference. Papers accepted in this session will describe innovative applications of logic programming to industrial problems. The application's innovativeness and industrial impact will be the main criteria used for judging the paper. Papers accepted for this session will be published in the proceedings as shorter, (up to) 10 pages papers. Posters ------- Posters provide a forum for presenting work in an informal and interactive setting. They are ideal for discussing current work not yet ready for publication, for PhD thesis summaries and research project overviews. Accepted posters will also get a 10 minute slot for presentation during the conference. Extended abstract (2 pages) of each accepted poster will be published in the proceedings. Posters must be submitted electronically. More information on the submission procedure will be available at http://www.utdallas.edu/ICLP05 Doctoral Student Consortium --------------------------- The Doctoral Consortium will provide an opportunity for students pursuing their doctoral thesis in logic programming and related areas to explore their research interests under the guidance of a panel of distinguished experts in the field. The Doctoral Consortium will also offer invited speakers and discussion groups. The Consortium will allow participants to interact with established researchers and with other students, through presentations, question-answer sessions, panel discussions, and invited presentations. A separate call-for-participation will be issued for the doctoral consortium. About 6 to 8 students will be selected for the consortium. Selected students will also present their research in the poster session. The abstract of the poster will be published in the conference proceedings. Financial support for selected students may be available. Publication ----------- The proceedings of the conference will be published by Springer-Verlag in the LNCS series. The proceedings will include the accepted papers and the abstracts of accepted posters. Sponsoring and prizes --------------------- The conference is sponsored by the Association for Logic Programming. The ALP has funds to assist financially disadvantaged participants. The ALP is planning to sponsor two prizes for ICLP'05: for the best technical paper and for the best application paper. Important dates --------------- POSTERS PAPERS Abstract submission deadline: 30 April Submission deadline: 1 June 6 May Notification of authors: 1 July 24 June Camera-ready copy due: 15 July 15 July Organization ------------ Conference Co-Chairs: Pedro Meseguer (IIIA-CSIC, Spain) Javier Larrosa (Technical University of Catalonia, Spain) Program Co-Chairs: Maurizio Gabbrielli (University of Bologna, Italy) Gopal Gupta (University of Texas at Dallas, USA) Workshop Chair: Hai-Feng Guo (University of Nebraska at Omaha, USA) Doctoral Consortium Chair: Enrico Pontelli (New Mexico State University, USA) Publicity Chair: Felip Manya (IIIA-CSIC, Spain) Program Committee: Roberto Bagnara University of Parma, Italy Maurice Bruynooghe KU Leuven, Belgium Giorgio Delzanno University of Genova, Italy Stefan Decker Digital Enterprise Research Institute, Ireland Thom Fruehwirth University of Ulm, Germany Maurizio Gabbrielli University of Bologna, Italy (Program Co-Chair) Gopal Gupta University of Texas at Dallas, USA (Program Co-Chair) Patricia Hill University of Leeds, UK Joxan Jaffar University of Singapore, Singapore Bharat Jayaraman SUNY Buffalo, USA Javier Larrosa Technical University of Catalonia, Spain (Conference Co-Chair) Michael Leuschel University of Southampton, UK Massimo Marchiori University of Venice, Italy and W3C, MIT, USA Pedro Meseguer IIIA-CSIC, Spain (Conference Co-Chair) Juan J. Moreno Navarro Technical University of Madrid, Spain Gopalan Nadathur University of Minnesota, USA Illka Niemela Helsinki U. of Tech. Finland Catuscia Palamidessi INRIA, France Enrico Pontelli New Mexico State Univeristy, USA I.V. Ramakrishnan SUNY Stony Brook, USA Vitor Santos Costa Federal U. of Rio de Janeiro, Brazil Harald Sondergaard University of Melbourne, Australia Peter Stuckey University of Melbourne, Australia Frank Valencia University of Uppsala, Sweden Contact addresses ----------------- Conference Chairs: Pedro Meseguer Javier Larrosa IIIA-CSIC Dep. LSI, UPC Campus UAB Jordi Girona 1-3 08193 Bellaterra, Spain 08034 Barcelona, Spain Program Co-chairs: iclp05-chairs(a)cs.unibo.it Maurizio Gabbrielli Gopal Gupta Department of Computer Science Department of Computer Sciences MS EC31 University of Bologna The University of Texas at Dallas Mura A. Zamboni 7 2601 N. Floyd Rd 40127 Bologna, Italy Richardson, TX 78050, USA

1 0

Job Advert
by N Ghani 20 Feb '05

20 Feb '05

EPSRC Funded Postdoctoral Fellowship Applications are invited for a research position available to work with Dr Neil Ghani and Dr Thorsten Altenkirch in the project THEORY AND APPLICATIONS OF CONTAINERS (EPSRC EP/C511964/1). The aim of this project is to develop both the theoretical foundations of containers and apply the results to other problems in Computer Science. The position is available from May 1 2005 and the appointment is for three years. Applicants should have a PhD in Computer Science or a related subject and a research record on developing and applying the theoretical foundations of computer science. Experience in category theory, functional programming and type theory are all desirable. Closing Date 15 March 2005 Contact: Dr Neil Ghani (ng13(a)mcs.le.ac.uk) and Dr Thorsten Altenkirch (txa(a)cs.nott.ac.uk)

1 0

Snapshots of Haskell software (update)
by Sven Panne 20 Feb '05

20 Feb '05

I've uploaded fresh bleeding edge binary snapshots of Alex, Haddock and Happy for Win32 and x86 Linux to http://aedion.de/haskell/ The corresponding sources are available there, too. Hugs currently has a few build problems, but a new snapshot of it will be available there as well soon. Feedback is, as always, appreciated... Have fun, S.

1 0

[ANNOUNCE] Non-ASCII Charset Support patch for Text.Regex
by shelarcy 19 Feb '05

19 Feb '05

Hello. Sakai Masahiro made a patch for GHC Text.Regex. Because Text.Regex doesn't support Non-ASII Charset well. He uses Onigurma what Regex C Library support Non-ASCII Charset well and GNU Regex Library replacement (this is usesd under Ruby version higher 1.9.X), replace for current Text.Regex's C library. Here is patch http://www.tom.sfc.keio.ac.jp/~sakai/d/?date=20050209#p01 and here is Onigurma's WebSite. http://www.geocities.jp/kosako3/oniguruma/ If you add this patch for GHC-6.2.2 sourcecode , autoconf, and ./configure with --enable-oniguruma option, then you can use Text.Regex with Oniguruma. -- shelarcy <shelarcy capella.freemail.ne.jp> http://page.freett.com/shelarcy/

1 0

class assosiated types, via GADTs.
by John Meacham 19 Feb '05

19 Feb '05

I believe there is a realationship between GADTs and class assosiated types which hints at a much simpler implementation of class assosiated types (CATs) and possibly a lessoning of the constraints mentioned in the paper. Assuming you have already implemented GADTs the translation is straightforward, which is why I think it might be relevant to ghc developers. An example is worth a thousand words so here it is: Take the example from the paper: > class MapKey k where > data Map k v > lookup :: k -> Map k v -> Maybe v > empty :: Map k v > > instance MapKey Int where > data Map k v = MapInt (IntMap v) > lookup k (MapInt m) = Data.IntMap.lookup k m > empty = MapInt Data.IntMap.empty > > instance MapKey () where > data Map k v = MapUnit (Maybe v) > lookup () (MapUnit x) = x > empty = MapUnit Nothing now here is the translation: > class MapKey k where > lookup :: k -> Map k v -> Maybe v > empty :: Map k v > > data MapKey k => Map k v where > MapUnit :: Maybe v -> Map () v > MapInt :: IntMap v -> Map Int v > > instance MapKey Int where > lookup k (MapInt m) = Data.IntMap.lookup k m > empty = MapInt Data.IntMap.empty > > instance MapKey () where > lookup () (MapUnit x) = x > empty = MapUnit Nothing The idea should be clear, each CAT data declaration is lifted to the top level. each instance declares a new constructor for said class with an explicit GADT type fixing the type which the instance is being declared for. I have not worked out any theory behind it, but it seems to work in manual translation testing and has a certain intuitiveness about it. One might wonder whether we are paying a price for turning Map into a multi-construtor type, but we are just moving an implicit branch on an unknown function in the dictonary into an explicit case. (which is a win in my books) Since a 'Map Int v' cannot come about except via the MapInt constructor, we are guarenteed that the pattern matching in the instance declarations cannot fail if the program typechecks. The only caveats I can think of to actually implementing this in ghc are separate compilation, the Map GADT must be marked somehow as 'open' since later instances might create new constructors for it. Hopefully this won't interfere with optimization too much. Perhaps some internally generated rules to turn lookup :: Int -> Map Int v -> Maybe v into lookupInt :: Int -> IntMap v -> Maybe v would suffice. (this transformation is valid because we know that if the type of Map is Map Int v then its constructor has to be MapInt, despite the 'open' nature of the Map datatype) Once a rule fires, we no longer have any reference to the open GADT so can optimize like normal. * The front end will still have to be modified to type CATs similarly to as described in the paper, if nothing else for sane user error messages. (however, I believe some of the GADT machinery may be reusable) After typechecking, the translation can be carried out even before (and independently of) the desugaring of typeclasses. * It might just not work with multiparameter type classes, I have not thought about it at all, but then again CATs obviate much of the need for multi-parameter type classes, so having them mutually exclusive (at first) doesn't seem like too much of a restriction. I may be completely wrong about this, or perhaps it is already well known. or perhaps it is what the paper actually says to do and I am just misreading it :) I thought I'd share because I have really wanted the class assosiated types in various situations and I wasn't sure if I'd have time (or knowledge) to explore this seriously too much further. John -- John Meacham - ⑆repetae.net⑆john⑈

5 11