From mark@felinemail.zzn.com Mon Sep 1 12:37:29 2003 From: mark@felinemail.zzn.com (Mark Espinoza) Date: Mon, 1 Sep 2003 06:37:29 -0500 Subject: Haskell 6.0 install Message-ID: <4B84EA44ED049C94C8E2BEABAB4DD5BA@mark.felinemail.zzn.com>

Hello, I let make run while I was sleeping on my redhat 7.3 system and in the morning it seemed to have worked. I ran make install which also seemed to work. Now when I type ghc --help it tells me about "ghc-5.04.2". Does this mean that I am still running ghc 5? What can I do if I am still running ghc 5? Will it cause me any serious problems if I start going through "An Introduction To Functional Programming Systems Using Haskell" and deal with it later? Thanks. Sincerely, Mark

Get your free e-mail address at http://felinemail.zzn.com
__________________
http://www.zzn.com

From droundy@abridgegame.org Mon Sep 1 13:08:55 2003 From: droundy@abridgegame.org (David Roundy) Date: Mon, 1 Sep 2003 08:08:55 -0400 Subject: Haskell 6.0 install In-Reply-To: <4B84EA44ED049C94C8E2BEABAB4DD5BA@mark.felinemail.zzn.com> References: <4B84EA44ED049C94C8E2BEABAB4DD5BA@mark.felinemail.zzn.com> Message-ID: <20030901120854.GB714@jdj5.mit.edu> On Mon, Sep 01, 2003 at 06:37:29AM -0500, Mark Espinoza wrote: > Hello, > > I let make run while I was sleeping on my redhat 7.3 system > and in the morning it seemed to have worked. I ran make install > which also seemed to work. Now when I type ghc --help it tells > me about "ghc-5.04.2". Does this mean that I am still running > ghc 5? What can I do if I am still running ghc 5? Will it > cause me any serious problems if I start going through "An > Introduction To Functional Programming Systems Using Haskell" > and deal with it later? Thanks. You can try running with ghc6 rather than ghc. ghc 6 and 5 can be installed in parallel without problems, and if you're writing pure haskell 98 with no extensions (as I imagine the book does), you should be able to use either one with no problems. -- David Roundy http://www.abridgegame.org From ganesh@earth.li Mon Sep 1 14:52:19 2003 From: ganesh@earth.li (Ganesh Sittampalam) Date: Mon, 01 Sep 2003 14:52:19 +0100 Subject: IO StateTransformer with an escape clause In-Reply-To: <200308261433.28038.thomas_bevan@toll.com.au> References: <200308261433.28038.thomas_bevan@toll.com.au> Message-ID: <5lj6lvg97rq98boudqj3idg6i9epmi8u32@4ax.com> On Tue, 26 Aug 2003 14:33:28 +1000, "Thomas L. Bevan" wrote: >Hi, > >I'd like some help building an IO StateTransformer which can be escaped midway >through without losing the state accummulated up to that point. >I tried constructing a > StateT s MaybeIO a monad but the state is lost when the rest of >the monad collapses. How is your MaybeIO type constructed? If with a monad transformer, then you could consider putting the MaybeT transformer outside the StateT transformer (I think that should work). Ganesh From m.p.donadio@ieee.org Wed Sep 3 00:54:51 2003 From: m.p.donadio@ieee.org (Matthew Donadio) Date: Tue, 02 Sep 2003 19:54:51 -0400 Subject: OT - Book on Programming Message-ID: <3F552DCB.BF6F31DA@comcast.net> Hi all, This is a little off-topic, but I know there are a lot of people here with teaching experience. I am looking for a good book on programming. I recently picked up a gig doing ActionScript programming for a multimedia company. They have very good designers, but not very good programmers. So, they decided to hire a professional programmer to help them out with some of their projects. They have several people in the company who understand the mechanics of programming, but don't really understand the concepts. I am looking for a book to recommend to them. I would really like to find a book that just discusses programming, and avoids any one particular language. It would have to cover the common imperative controls, as well as basic data structures. Thanks. -- Matthew Donadio (m.p.donadio@ieee.org) From dvanhorn@cs.uvm.edu Wed Sep 3 01:19:02 2003 From: dvanhorn@cs.uvm.edu (David Van Horn) Date: Tue, 02 Sep 2003 19:19:02 -0500 Subject: OT - Book on Programming In-Reply-To: <3F552DCB.BF6F31DA@comcast.net> References: <3F552DCB.BF6F31DA@comcast.net> Message-ID: <3F553376.1070500@cs.uvm.edu> How to Design Programs: An Introduction to Programming and Computing by Felleisen, Findler, Flatt, and Krishnamurthi is rather good at introducing the fundamentals of programming in my opinion. I've given the book to friends of mine with no background in programming or computer science and they've found it quite accessible. The text is freely available on-line too. http://htdp.org/ The book is about the general design of programs, but uses Scheme as a teaching language, and the DrScheme environment (also freely available) is tailored for use with the book. If your audience has any background in programming, they should be able to move through the text quickly. HTH -d From Tom.Pledger@peace.com Wed Sep 3 04:35:03 2003 From: Tom.Pledger@peace.com (Tom Pledger) Date: Wed, 3 Sep 2003 15:35:03 +1200 Subject: Poll result: How to respond to homework questions In-Reply-To: <16205.9696.324126.137830@tux-17.corp.peace.com> References: <16205.9696.324126.137830@tux-17.corp.peace.com> Message-ID: <16213.24935.818437.201078@tux-17.corp.peace.com> Hi. The stream of votes has dried up, and the ICFP people and monthly digest people have had an opportunity, so here's the collated result. 22 people voted. (A) Give a perfect answer. (B) Give a subtly flawed answer. (C) Give an obfuscated answer. (D) Give a critique of what the questioner has tried so far, or ask to see what the questioner has tried so far. (E) Give relevant general advice without answering the specific question. Approve Neither Disapprove A 1 6 15 B 1 4 17 C 3 4 15 D 20 2 0 E 21 1 0 Combination Count +ADE -BC 1 +BDE -A 1 +C 1 +CDE -AB 2 +DE 3 +DE -ABC 11 +DE -BC 2 +E -ABC 1 Notes ----- Some votes were made conditional on the tone and content of the question. I smoothed these out for ease of collation, e.g. "+ADE -BC if ..., otherwise +DE -ABC" became +DE -BC. 15 of the votes were accompanied by some descriptive text. There's not much I can do to tabulate that extra information, but here are a couple of the things I thought stood out. - Some of the 11 like-minded people showed clearly that they're university staff: they gave longer forms of their email signatures than they normally use for posting to lists @haskell.org. - "D and E are what I do when a student fronts up to my office wanting help with homework I've set. I'd have no problem with (informed) third parties doing the same." (If you sent me descriptive text, didn't keep a copy yourself, and were expecting me to relay it to everyone, please let me know.) - Tom From ajb@spamcop.net Wed Sep 3 06:52:10 2003 From: ajb@spamcop.net (Andrew J Bromage) Date: Wed, 3 Sep 2003 15:52:10 +1000 Subject: Help on Homework help Message-ID: <20030903055210.GA27774@smtp.alicorna.com> G'day all. In response to the recent poll, and because I promised to, here is a first attempt at a page to direct people to, should they ask for homework help in the wrong way: http://haskell.org/hawiki/HomeworkHelp Comments, criticisms and contributions are most welcome. Cheers, Andrew Bromage From simonpj@microsoft.com Wed Sep 3 08:34:58 2003 From: simonpj@microsoft.com (Simon Peyton-Jones) Date: Wed, 3 Sep 2003 08:34:58 +0100 Subject: Help on Homework help Message-ID: <4B93206CA3C55D4F96A61C9B1DC80DE33B5A8F@EUR-MSG-03.europe.corp.microsoft.com> Good write-up. One suggestion: it'd be good to suggest *strongly* that people only send = their homework-style questions to Haskell-caf=E9, not to the main = Haskell list. Given the existence of the fomer, using the latter is = likely to irritate, which isn't likely to elicit helpful responses. I know I could make the change myself, but you'll do it better :-) Simon | -----Original Message----- | From: haskell-cafe-admin@haskell.org = [mailto:haskell-cafe-admin@haskell.org] On Behalf Of Andrew | J Bromage | Sent: 03 September 2003 06:52 | To: haskell-cafe@haskell.org | Subject: Help on Homework help |=20 | G'day all. |=20 | In response to the recent poll, and because I promised to, here is a | first attempt at a page to direct people to, should they ask for | homework help in the wrong way: |=20 | http://haskell.org/hawiki/HomeworkHelp |=20 | Comments, criticisms and contributions are most welcome. |=20 | Cheers, | Andrew Bromage | _______________________________________________ | Haskell-Cafe mailing list | Haskell-Cafe@haskell.org | http://www.haskell.org/mailman/listinfo/haskell-cafe From ajb@spamcop.net Wed Sep 3 08:44:03 2003 From: ajb@spamcop.net (Andrew J Bromage) Date: Wed, 3 Sep 2003 17:44:03 +1000 Subject: Help on Homework help In-Reply-To: <4B93206CA3C55D4F96A61C9B1DC80DE33B5A8F@EUR-MSG-03.europe.corp.microsoft.com> References: <4B93206CA3C55D4F96A61C9B1DC80DE33B5A8F@EUR-MSG-03.europe.corp.microsoft.com> Message-ID: <20030903074403.GA29516@smtp.alicorna.com> G'day all. On Wed, Sep 03, 2003 at 08:34:58AM +0100, Simon Peyton-Jones wrote: > One suggestion: it'd be good to suggest *strongly* that people only > send their homework-style questions to Haskell-café, not to the main > Haskell list. Done, thanks. Cheers, Andrew Bromage From shae@ScannedInAvian.com Wed Sep 3 08:44:51 2003 From: shae@ScannedInAvian.com (Shae Matijs Erisson) Date: Wed, 03 Sep 2003 09:44:51 +0200 Subject: OT - Book on Programming In-Reply-To: <3F552DCB.BF6F31DA@comcast.net> (Matthew Donadio's message of "Tue, 02 Sep 2003 19:54:51 -0400") References: <3F552DCB.BF6F31DA@comcast.net> Message-ID: <87r82y71u4.fsf@thunderbird.webwitches.com> Matthew Donadio writes: > They have several people in the company who understand the mechanics of > programming, but don't really understand the concepts. I am looking for > a book to recommend to them. I would really like to find a book that > just discusses programming, and avoids any one particular language. It > would have to cover the common imperative controls, as well as basic > data structures. I'd also recommend "The Pragmatic Programmer" http://www.pragmaticprogrammer.com/ppbook/index.shtml It's difficult to summarize, but I think every programmer should read it at least once. It is mostly about the mechanics of programming also. -- Shae Erisson - this space for rent - #haskell on irc.freenode.net "Notwithstanding fervent argument that patent protection is essential for the growth of the software industry, commentators have noted that `this industry is growing by leaps and bounds without it.'" -- US Supreme Court Justice John Paul Stevens, March 3, 1981. From wvasconc@csd.abdn.ac.uk Wed Sep 3 09:30:18 2003 From: wvasconc@csd.abdn.ac.uk (Wamberto Vasconcelos) Date: Wed, 3 Sep 2003 09:30:18 +0100 Subject: Problem with Infinite Lists Message-ID: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> Folks I am new in this forum, so apologies if this has been asked before. If this is the case, please point me to the direction of the answer and I won't bother you again! Also, could you make sure you CC your answers to me? I am not sure how long it takes for my subscription to be activated and I don't want to miss out on the action :-) The Problem I have defined this function to calculate the Fibonacci numbers: all_fib :: [Float] all_fib = 1:(1:(add_fib all_fib)) where add_fib (x:y:rs) = (x + y):(add_fib (y:rs)) Which seems to work: Main> take 20 all_fib [1.0,1.0,2.0,3.0,5.0,8.0,13.0,21.0,34.0,55.0,89.0,144.0,233.0,377.0, 610.0,987.0,1597.0,2584.0,4181.0,6765.0] However, when I tried Main> filter even (take 20 all_fib) ERROR - Illegal Haskell 98 class constraint in inferred type *** Expression : filter even (take 20 all_fib) *** Type : Integral Float => [Float] What is going on here? Thanks in advance for any help/hint. -- Wamberto Vasconcelos, PhD wvasconcelos@acm.org Department of Computing Science http://www.csd.abdn.ac.uk/~wvasconc University of Aberdeen, Aberdeen AB24 3UE, Scotland, UK Phone +44 (0)1224 272283 Fax +44 (0)1224 273422 From hinsen@cnrs-orleans.fr Wed Sep 3 11:51:21 2003 From: hinsen@cnrs-orleans.fr (Konrad Hinsen) Date: Wed, 3 Sep 2003 12:51:21 +0200 Subject: Problem with Infinite Lists In-Reply-To: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> References: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> Message-ID: <200309031251.21245.hinsen@cnrs-orleans.fr> On Wednesday 03 September 2003 10:30, Wamberto Vasconcelos wrote: > Which seems to work: > > Main> take 20 all_fib > [1.0,1.0,2.0,3.0,5.0,8.0,13.0,21.0,34.0,55.0,89.0,144.0,233.0,377.0, > 610.0,987.0,1597.0,2584.0,4181.0,6765.0] > > However, when I tried > > Main> filter even (take 20 all_fib) > ERROR - Illegal Haskell 98 class constraint in inferred type > *** Expression : filter even (take 20 all_fib) > *** Type : Integral Float =3D> [Float] > > What is going on here? "even" wants an integral type for its argument, but you are applying it t= o a=20 list of floats. Konrad. From steffen.mazanek@unibw-muenchen.de Wed Sep 3 11:50:38 2003 From: steffen.mazanek@unibw-muenchen.de (Steffen Mazanek) Date: Wed, 03 Sep 2003 12:50:38 +0200 Subject: Problem with Infinite Lists In-Reply-To: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> References: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> Message-ID: <3F55C77E.8050603@unibw-muenchen.de> Hello, >all_fib :: [Float] > > > You define "all_fib" to return a list of "Float", but "even" does only work for numbers whose type is an instance of the class Integral, e.g. Int. HTH and ciao, Steffen From gwright@comcast.net Wed Sep 3 14:11:28 2003 From: gwright@comcast.net (Gregory Wright) Date: Wed, 3 Sep 2003 09:11:28 -0400 Subject: OT - Book on Programming In-Reply-To: <3F552DCB.BF6F31DA@comcast.net> Message-ID: <21170FAA-DE10-11D7-84AE-00039398F084@comcast.net> On Tuesday, September 2, 2003, at 07:54 PM, Matthew Donadio wrote: > > They have several people in the company who understand the mechanics of > programming, but don't really understand the concepts. I am looking > for > a book to recommend to them. I would really like to find a book that > just discusses programming, and avoids any one particular language. It > would have to cover the common imperative controls, as well as basic > data structures. > In my previous company, I recommended Thomas Standish's _Data Structure Techniques_. It's out of print, but easily (and inexpensively) available at Amazon. I like the original 1980 edition; his later _Data Structure Techniques in Java_ is too closely tied to the Java language. It covers the usual data structures (lists, trees, strings, etc.) using pseudocode algorithms. It's at the right level for someone who has gotten the hang of programming at the level of compiling and debugging simple programs and is ready to move onto something more complicated. Best Wishes, Greg Gregory Wright Antiope Associates 18 Clay Street Fair Haven, New Jersey 07704 USA gwright@antiope.com From hinsen@cnrs-orleans.fr Wed Sep 3 14:13:30 2003 From: hinsen@cnrs-orleans.fr (Konrad Hinsen) Date: Wed, 3 Sep 2003 15:13:30 +0200 Subject: Histograms Message-ID: <200309031513.30989.hinsen@cnrs-orleans.fr> How would one best (read: most efficiently) create a histogram in Haskell= ?=20 More precisely, given a list of integers in a known range, I want a count= of=20 how often any given number occurs. In imperative languages this would be done by iterating over the list and= =20 updating the count for the element under consideration. Without updatable= =20 data structures, the most straightforward solution I can think of would s= tart=20 by sorting the list, but for a long list that is very inefficient. Any suggestions? Konrad. From wferi@afavant.elte.hu Wed Sep 3 14:51:25 2003 From: wferi@afavant.elte.hu (Ferenc Wagner) Date: Wed, 03 Sep 2003 15:51:25 +0200 Subject: Histograms In-Reply-To: <200309031513.30989.hinsen@cnrs-orleans.fr> (Konrad Hinsen's message of "Wed, 3 Sep 2003 15:13:30 +0200") References: <200309031513.30989.hinsen@cnrs-orleans.fr> Message-ID: <3o65ka9e02.fsf@tba.elte.hu> Konrad Hinsen writes: > How would one best (read: most efficiently) create a > histogram in Haskell? Check out accumArray. Feri. From henning@ikso.net Wed Sep 3 15:28:18 2003 From: henning@ikso.net (henning@ikso.net) Date: Wed, 3 Sep 2003 14:28:18 -0000 (UTC) Subject: haskell html-colorizer Message-ID: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> Are anyone aware of a haskell html-colorizer a' la the emacs-mode? Regards /Henning From droundy@jdj5.mit.edu Wed Sep 3 15:39:40 2003 From: droundy@jdj5.mit.edu (David Roundy) Date: Wed, 3 Sep 2003 10:39:40 -0400 Subject: haskell html-colorizer In-Reply-To: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> References: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> Message-ID: <20030903143938.GX15540@jdj5.mit.edu> On Wed, Sep 03, 2003 at 02:28:18PM -0000, henning@ikso.net wrote: > Are anyone aware of a haskell html-colorizer a' la the emacs-mode? I've never used it, but webcpp may be what you're looking for. http://webcpp.sourceforge.net/ -- David Roundy http://civet.berkeley.edu/droundy/ From Malcolm.Wallace@cs.york.ac.uk Wed Sep 3 15:49:47 2003 From: Malcolm.Wallace@cs.york.ac.uk (Malcolm Wallace) Date: Wed, 3 Sep 2003 15:49:47 +0100 Subject: haskell html-colorizer In-Reply-To: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> References: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> Message-ID: <20030903154947.08c84e68.Malcolm.Wallace@cs.york.ac.uk> This is a multi-part message in MIME format. --Multipart_Wed__3_Sep_2003_15:49:47_+0100_08397710 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit writes: > Are anyone aware of a haskell html-colorizer a' la the emacs-mode? I wrote one a few months ago, but never released it. So here it is, released at last as version 1.0. (It can also do ANSI terminal colouring, if you are interested.) Regards, Malcolm --Multipart_Wed__3_Sep_2003_15:49:47_+0100_08397710 Content-Type: application/octet-stream; name="hscolour-1.0.tar.gz" Content-Disposition: attachment; filename="hscolour-1.0.tar.gz" Content-Transfer-Encoding: base64 H4sIAAv/VT8AA+07/VMbO5L5dfVX6ObthfFhO+YrvPOL2UcICewjIRWSTaWABeGRbRXjGTOaCfhC 9m+/7pY0XzYhW8tL3e2zqszMqFvdra/+khjpfhzGWdJaaXeePPp9Sqez3tnc2IBnZ3VlcxWfnc6K fZryqLO5ubqxtrnZWV95BLCN1c1HfON3kqdSMp2KhPNHYxHCSIzvxLseSRn+CIF+bBmV57/tvh6W R2el03m6vn7n/G+ub9Tmf31tc+UR7zysGPPLH3z+d2i+3yZyoBnnX/ilnF7HScB5jx9/iAKZwIpI ZPMl/BkmcRYF/FUiZXQKyE1EHobTyQiRSxjvZGDgoZjGWcp5Db4zFZZAPx6PZZTWEZ6HmXQIkQJh COFNnMAkmfrPIplbD/jxxNY/j8OgOU8saJwjzRVLp4mKhjMIr8UQhBWW0wiXzbdxomx8IZNv42gZ yn6q4igXmd+Fip0WUVpp8UKNKw2gk01eTBw2/Mq+Mf+V/b/95mgfdMBDr7F79j9s+7Xa/l97urq6 2P8/orRa/K1IUiXCcMpTcSkjPkjiMd/LhppvR1od9XG/w6rgobpIRDLtsnEcZKHkuFpgefl8pIaj EH7pYUTrtPgeDKoVZueEUiQyDu3bRf4Wqkja1xfxdWRfP0zwRVPTYZzGhkaW6NiA6M3i0/uBHKTu /V3OVYvP8m2sFW6cJk+kTmF7uArC2HNS+u12g2qMcsw/zU7ZmYJY9C0jcRFKGKA4DN/JIRHW9IWC mTcUzOhC6NzHREyY6YhM38tkrCIRHqn/kX+Caeh2+f4h9/ejtAm/BgzstQrSUXMkUSaUoMFhESaS MTWexEnKD5ROua+iFDb7BH6yqTRqcnVzOGg4pHdZBKSSLNqOgndSBEdpADo5B++gGvOVfqGGqlRL Pc0HhDEGAj4XWvWxW7AeUNGm0D94BlJ3GUunE8lhOHlRekVnGBOpq4Z+Il5rix8ZNZu/MZzeHAna Icg+LMYoHsscw9bB4uD1OuTn3zSnDa5RECJ8w6d8eZlrln+RkEsnu0c7S92l46Wur0fxNWH5S78s denrBtt4e16jUWKe01zhK4y5xWjqPaR3vO2xYmUW1c9dNS3MvHrHVePaLbBfeEDc7pYy8d885rZO uXrF1eNaK9WvunojS17/V1ttJC/IQD1M984ojrWEJaf6I/kZzFg84OlI8gGs6vgaJw7ZaK4015lK cSsALOFg8hOupzqV425pbkri/NWjb+CB+G93+LVKR6RO2kefjoB3vdXfD7z827Y6gmUNQgTxNWPl rV1itOmx2jZ3kJ89s6IDkQq71i1ps/5lYJRgfRu4JnnN/U2YisDARH3Jd8EbKLU0W5lTKwIZNybv dacMQ7+gNCIrZRh4AOU9t1qGFY4AwdYrNGH6Lot2G2XYO5xwLf+mAhkDbLMM24mhNyKU1v36uQzz S55IvwHQtQ5fLsD9Cu5z0b8s467XcFluNnBjH+dDdzpXdxTIIk0TWJNAsWSYbC3uf/xTMVGsjHd8 agekUul8zFmK+bL2GJ9XgBnoyr6oamru/eJxfywmnHRO23W7Yag2GncR88YeKwufs++M7aLejriK BipSqYSNOZmAXYeda2yXbrOSEaNhNQv2tFLf4316HqM3ic540/qhTXL+m8ZTJm7G+uEEzFrDO7X4 HFT0wlKwqEExnsvLpIAJsrycK2TAgWFIoLeFhc1nATSms8AlTfDaI1kPQF+1rsEGT3DpgBcjmbPK /H2SyWIuN0deAXopQi0LUGiIvQTlw8lCcwFvxkqTjrR23VpKVjP1dTs/A+/xIKbZhyiEP2vxuunm Hmg1RGxlZsmFMuUhLnejj6EZVScyzZKI+wMU1Ns5PPjw+s2RB5hNbqoO9t/sUgUuNqeOuAHecLsR jCLKa33d1brBb+H1vHA2zmm3JSAisgthBLj1KLiet5JveQyWJLlWumRPLQetTef5OboYAroU7Efn KE3PzMRc8M2v/ln3RqMa0VXJoCe3t7P4mn0rKvrjlEr8VzNeDxUK3pf/2VhbrcV/T1c21hbx348o Nparuy0Y1tWin3p4lEcjZS+qB24FmHXY4WA64C/ZC3h+kug0wgtaE3hYewJvaEvg8XEE9gqogr+i PqN69nevmkeg75uo+ZpkGy2rwofqkS4wtpmhWkE/iV7AKaJn4QAZOPo89Fb2cKgid2voq5wWop6Z 5oXHktfOF7jx/0W9zNn/oJUfNgl0z/6HTV/P/2501jYW+/9HlMr+V2RdfX6nUnApEZcvbpLRr9f1 S8TqOYs6TVt9RPEi98EZ2o0+V7IT5WxHRdkQQ9A4d6Swi/S0S0Tn+eYitWwzyTZxbFPDpRywyfO6 TG41YTsnIVuNUzD3erd+CORAZGH6XX2pZ5CNG15N8/5fT8fXktSoqv+g6XhW2zQ2JCjPfR0DIjKR 9ke4O4MYOIcUGQDSSwUb1csP7lwQ6px/8skBnQJK/0RCDOYIGVKU0wJaZuNxb+/w9W4RydYZ+YgO 0VdxUugV/v1cno6rBc4uesBiucLAkSh3uxLoH/vmvVnaYo3TUuMJtIGRoqj6RLe2fN3sh0JrNZga GDDjbT5EF6TN0/hSRqikKDSUN3xP6EtwU7gGcn1JmU2IpNOYC4OL61CKMXMNi3QjSWdeTwuwC6CO S3V+v0vhk9JHE4FM6tFRj1B0o5vLR7lqo0Uh/NJN/KQ4ZwKOk6OjCxba0oGKO4hAZU5lREEbjHaY BbIdwjBAxIYj8gpHScurTIJfpCmwRVraDMkYU0qZloMs5BdhfKGhCY2rf9X12l436hoGt6TC20p/ mExg6/nE7qqB+e6rTIRqoMBNjMRYIoF5GY+emS7fv4Jl1/aWlyMYGqLNDMA79yyv3p+gAJVz0kbn OClOnfcFDAHh0zjEA7Myr6Ju0RzQkdjy8lUErM494JO3IPSzimAWvSqMiY5dkCtgMfm93lJrqcE1 iSYpY+5s0Gxffb28bLNF/Rn+1WLnsp/P5AUszkvuL51ES+egksbnDYyl7SB98bpeyzPS9fC7BR2t sEDxInjCfNwt4QxTbOGsRqfgVx4HWKmWjQPbFJsDw/5gZTpF1sjtKRpst/ltXaPSJqr28ZZHW70O 9tR1VTeayfiOrNrcLmrEr/XQj5ZXaFTrrIGt97Vg3eGGdevrA7JufQ/rnuk1cm5CVR27uj7zMdIP JWU0KyElbmSSxAn3opj3w1ijxnQIIChjX1rGr3uPGuY9niRhZPebnH5E3+cW316RY3PLD4wPg/Ga IUCR2z5i/U0k9ITvw4n5xidQslraaiN4vDGewC3fRcFqUVyDtb6SCtxxxsNoPpYbk29aqZL/V7M+ v/qj7pkZ4sICjKqjW3GgZpRIGc+NINK2uF6r5dXycnZCHz/mIpoW5mIuFVIBccpBqYdTgCQJODaO 9JcZ0neKoblRQM571Y6Z82bvwCXnVRe4xpmdh2w8WfT5NCFbz7aE6izOzPAaF7aCeRBfz8M0Tm0J c+TY6+n4Ig51GRM82DImTFh3aWbr9KxDXMM8n4c5l6Y3D9O6xzXUk6UZ3J7xkit9N1namb5bV7mE OpOxzdnn/i9h48ZxnpSbf5YvBNzXxx6aY6/p0faAJ+59fBjfkN7MbsTXGP7IEPDR5/bUAD5NQIgv Ef2BBemeoXtJ6MUcvcFrKBHfxLWGUiSv8cwaamOkCb1DYraKdBw8z+A3AFdHDRHY78NmhKfQ3ikr 1qvpU7sNkG4X/vTgd3ICf27h96wFf1pb8OdX+P0DwfhxDL9ToFJeyEiHfNfu8WmDe798+eo3mh5z iw3B3n/89Of/fPxfy+0nz3pbf/n15OTvt61/eP9KqqmS/9nTRq098B2ge/I/q5tPV+r3/56uri/y Pz+i2OQK3eWppGgwMqokZxgbCxW546kJxoUG0IbY8W0SD9+AGw8QUYVsJ0M0QmQBbShZTXKQby6c T172DcGajmQIStBAvNbIO70LAl8EnIGk6dS0AogLc+ggVvTBqwIoidbICaXj0BGaRUdoGV+cloWF cDqPlAXf2urxSZaCZ4DBpuVTlqqJrf+5ZiQdtvt2s1zMmbilPDzFkZ9jYy0Uhh9tUkWJxNTF+/ef 4LsaZ1O7nA/5nM/gbQscXoyxfQihnj2hilmfsj2Pyd771wd3cCGBcy/S9z5oMZRdDqwmwIYf42je 0tic8uMBjAfor1PgywrxwWOblz0onUkXuL5ujhrlCwBgHrUjRmJ+LzVCduQGcZSiLzjivtR9AU6u jbT3cBR1mg0GzOHALnB+kM4rwXsc6QYdtWIVEsppaiJG1ZV7JJzQDY3qLRID8Z5d4JRpmi54NYiV KyUOUY4LTHw3qJUbJg41KzAzh1i5bpIzx8qSAObTNKjcQan0o3IDpQKp3T8xTAiC6yrpuTsFfWBX sEUEx7V2K6VE4WL4HTSYnVp/6dlSt0/n0d7jMP0F8CykrwucrQJneBfO4wJHjCfzkfqEQru3y2fA uUqlULum/yv2/10W/Q7Xf++1/+sz9n/tKf7/x8L+//7FHvXAzNPJT/2yCSW60/EEteqc05wXCkPE OJli1nccf5ZojRr1Qx1zGbC5e6PSnTiQeH4MyuonNQCPn5+d7T3fOTtzbT5EWgzkW5mA2z3ePyxh bR/9tntw8N8/F7ivwO9QATPuR19qvf8C84Wm9icMGuowm4VeW9lYt/awxUVyoVK82OwOeX4C3a0G +cMJ8GavJCZ8tPcPd2+goeZ+VhW6wWzMYO8GcgocuAfCTFTg8YpQJve/H6UzDF8dbB+9Ovzoel4w J8YPzZZyvjICRydFfwImDGjgqODUY4aWZ5G6yiRPY56OlAb7TLTazC6Pajrc2kEHS+IYb7vWJOZ/ Ri9mYs8fcuG+IwlVnDZ4T4DHE1CLyMLmiEk/T4x1xfvPgusRxqOYq8DLWviDzY5xK1cwkrDQwXdq s5mrVpUugbzOutfx+uOguLeF97EMRczC2wHwRmNxKcHvGaihOV8BbwY7ru0mARLozKBFMY2N30Te MaJa/xgLbqSXQoUZGN0zlEzFu8Y1yrRM7Ku3YzvrYw6SiDf4AFrJwCtdL6T2biwb9m5V+cjHCGNv lLkt7vpnruNixbWC8cXk0YW0eAHMdqrCghKm/vC+sCNmeOKCwMuK4MAgMUxmTDhsK4WXCjlE6XSZ WQQBELyYmomsOK85gfzm5vFpDYA58dN5gBt3RHNTg/g39jbZDe8Wtf+Gt8aq9n93+8Xr3QfncZ/9 X9tYz+3/2uY62v/11cX///2Q4lI+Xb5dOwBt0QFoHo8lbdCm313YTjyZJhgWdTlf7XTWmvy1GWD+ EYyS6Es8Flfo4auULih/ipNLxp5nKgxg43dN5Ic6kzsRGTOBH4Fc5R3RH2PvRxhCxsNEjEGJRqXA UtRPekmB4T8VSIVJzvwGL7Nazo8nlN9ETg0OeBTX1YEoQaPNOTK2qUwGZhLp5leCTZum+acKPD/F 67qgpEUSUAqm+O+aQFJThv9xkd/dZvsD0LFGXpEMM0q8A48hDGTUBM1KyrZEU0Vo1zg/NLIqzUR4 LaaaXycqTYERGPMc2VlB9inO7JChtUIrU5KCOgQOwwD0L3AP5Y0CN4PDu0oVCH0xZTKAV1DeIqcg aIhIbvRJQJXnFwhARkM+S4hg4FzKNvO3Iy5vIOqBZgoPnvuYjglci0ChsbjIkHa7YQbeTeVY0MAI 6DDXiii4Sf8s8AwSz7PxoKmUi6LrKkAy6+MhaEbHVOj4BGygZBgAXGj73y/4H0+c33sX6btvI917 H+kbN5L+6TtJdAWlIv2/em3TUavfzCzdzSzdzizuZ87c0Czf0ZxzS7N+T/Oum5rz72r+u5nuRVmU RVmURVmURVmURVmURVmURVmU7y7/C2X2/1QAUAAA --Multipart_Wed__3_Sep_2003_15:49:47_+0100_08397710-- From chak@cse.unsw.edu.au Wed Sep 3 18:30:45 2003 From: chak@cse.unsw.edu.au (Manuel M T Chakravarty) Date: Wed, 03 Sep 2003 19:30:45 +0200 (CEST) Subject: haskell html-colorizer In-Reply-To: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> References: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> Message-ID: <20030903.193045.925163789.chak@cse.unsw.edu.au> > Are anyone aware of a haskell html-colorizer a' la the emacs-mode? XEmacs has the very nice function M-x htmlize-buffer which when executed in a buffer containing a Haskell file (or any other source code for which (X)Emacs provides special markup) produces HTML output that exactly matches the display used in (X)Emacs itself (ie, including colours). I am unsure whether that or similar functionality is provided by GNU Emacs, too, but it may very well be. Cheers, Manuel From dvanhorn@emba.uvm.edu Wed Sep 3 18:45:19 2003 From: dvanhorn@emba.uvm.edu (David Van Horn) Date: Wed, 03 Sep 2003 13:45:19 -0400 Subject: haskell html-colorizer References: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> <20030903.193045.925163789.chak@cse.unsw.edu.au> Message-ID: <3F5628AF.8070503@emba.uvm.edu> Manuel M T Chakravarty wrote: > XEmacs has the very nice function > > M-x htmlize-buffer > > which when executed in a buffer containing a Haskell file > (or any other source code for which (X)Emacs provides > special markup) produces HTML output that exactly matches > the display used in (X)Emacs itself (ie, including colours). > > I am unsure whether that or similar functionality is > provided by GNU Emacs, too, but it may very well be. I don't believe it is provided with GNU Emacs, but it can be added on: http://fly.cc.fer.hr/~hniksic/emacs/htmlize.el -d From john@repetae.net Wed Sep 3 18:50:46 2003 From: john@repetae.net (John Meacham) Date: Wed, 3 Sep 2003 10:50:46 -0700 Subject: haskell html-colorizer In-Reply-To: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> References: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> Message-ID: <20030903175046.GD25669@momenergy.repetae.net> On Wed, Sep 03, 2003 at 02:28:18PM -0000, henning@ikso.net wrote: > Are anyone aware of a haskell html-colorizer a' la the emacs-mode? typing :runtime! syntax/2html.vim in gvim will cause it to spit out an html file with the exact same syntax highlighting as used in the editor. quite handy. some versions of nenscript can spit out highlighted html too, not sure if they support haskell though. John -- --------------------------------------------------------------------------- John Meacham - California Institute of Technology, Alum. - john@foo.net --------------------------------------------------------------------------- From obraun@unsane.org Wed Sep 3 18:51:09 2003 From: obraun@unsane.org (Oliver Braun) Date: Wed, 3 Sep 2003 19:51:09 +0200 Subject: haskell html-colorizer In-Reply-To: <20030903.193045.925163789.chak@cse.unsw.edu.au> References: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> <20030903.193045.925163789.chak@cse.unsw.edu.au> Message-ID: <20030903175109.GA46744@unsane.de> --VbJkn9YxBvnuCH5J Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable * Manuel M T Chakravarty [2003-09-03 19:30 +0200]: > > Are anyone aware of a haskell html-colorizer a' la the emacs-mode? > XEmacs has the very nice function > M-x htmlize-buffer > which when executed in a buffer containing a Haskell file > (or any other source code for which (X)Emacs provides > special markup) produces HTML output that exactly matches > the display used in (X)Emacs itself (ie, including colours). Vim[1] has the same functionality. Just type :runtime! syntax/2html.vim in normal mode. Regards, Olli 1. http://vim.org/ --=20 Oliver Braun -- obraun @ { unsane.org | FreeBSD.org | haskell.org } --VbJkn9YxBvnuCH5J Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (FreeBSD) iD8DBQE/VioNwLFrfe8lsboRAjUeAKCS4P48kpmBh5hqPlQgNKPNtvynIACdHQAt XWPNAwBm+nWbQlbhzsy2ZzY= =dVYt -----END PGP SIGNATURE----- --VbJkn9YxBvnuCH5J-- From knarf@bka-inc.com Wed Sep 3 19:18:00 2003 From: knarf@bka-inc.com (Frank Seaton Taylor) Date: Wed, 3 Sep 2003 14:18:00 -0400 Subject: Problem with Infinite Lists In-Reply-To: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> Message-ID: This is a bit off topic, but... Warning: contains evangelism from a number theorist. The Fibonacci sequence should start with 0 and 1 rather than 1 and 1. Doing so makes it adhere to the following property: all_fib !! (gcd m n) == gcd (all_fib !! m) (all_fib !! n) for m, n nonnegative integers. With the exception that Haskell misdefines gcd 0 0 as an error rather than 0. ---Frank On Wednesday, Sep 3, 2003, at 04:30 US/Eastern, Wamberto Vasconcelos wrote: > Folks > > I am new in this forum, so apologies if this has been asked before. If > this is > the case, please point me to the direction of the answer and I won't > bother you > again! > > Also, could you make sure you CC your answers to me? I am not sure how > long it > takes for my subscription to be activated and I don't want to miss out > on the > action :-) > > The Problem > > I have defined this function to calculate the Fibonacci numbers: > > all_fib :: [Float] > > all_fib = 1:(1:(add_fib all_fib)) > where add_fib (x:y:rs) = (x + y):(add_fib (y:rs)) > > Which seems to work: > > Main> take 20 all_fib > [1.0,1.0,2.0,3.0,5.0,8.0,13.0,21.0,34.0,55.0,89.0,144.0,233.0,377.0, > 610.0,987.0,1597.0,2584.0,4181.0,6765.0] > > However, when I tried > > Main> filter even (take 20 all_fib) > ERROR - Illegal Haskell 98 class constraint in inferred type > *** Expression : filter even (take 20 all_fib) > *** Type : Integral Float => [Float] > > What is going on here? > > Thanks in advance for any help/hint. > > -- > Wamberto Vasconcelos, PhD wvasconcelos@acm.org > Department of Computing Science http://www.csd.abdn.ac.uk/~wvasconc > University of Aberdeen, Aberdeen AB24 3UE, Scotland, UK > Phone +44 (0)1224 272283 Fax +44 (0)1224 273422 > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > From wolfgang@jeltsch.net Wed Sep 3 20:16:53 2003 From: wolfgang@jeltsch.net (Wolfgang Jeltsch) Date: Wed, 3 Sep 2003 21:16:53 +0200 Subject: Problem with Infinite Lists In-Reply-To: <3F55C77E.8050603@unibw-muenchen.de> References: <1062577818.3f55a69aeb66a@www.csd.abdn.ac.uk> <3F55C77E.8050603@unibw-muenchen.de> Message-ID: <200309032116.53024.wolfgang@jeltsch.net> Am Mittwoch, 3. September 2003 12:50 schrieb Steffen Mazanek: > Hello, > > >all_fib :: [Float] > > You define "all_fib" to return a list of "Float", but "even" does only work > for numbers whose type is an instance of the class Integral, e.g. Int. > > HTH and ciao, > Steffen And I see no reason to define it as a list of Float because all Fibonacci numbers are natural numbers. So better write: all_fib :: [Integer] Wolfgang From nick.name@inwind.it Thu Sep 4 00:46:15 2003 From: nick.name@inwind.it (Nick Name) Date: Thu, 4 Sep 2003 01:46:15 +0200 Subject: Why not haskell? Message-ID: <200309040146.15981.nick.name@inwind.it> Inspired by the various replies to the "haskell for non-haskell sake" I wonder why so much people uses haskell "only for prototipying" or producing code in other languages. I am just curios to hear from people who do not use haskell for project releases, or just think it's not suitable for a mature project, what exactly they find bad in current haskell implementations, or perhaps in the standard. Vincenzo From db@zigo.dhs.org Thu Sep 4 05:27:51 2003 From: db@zigo.dhs.org (Dennis Bjorklund) Date: Thu, 4 Sep 2003 06:27:51 +0200 (CEST) Subject: haskell html-colorizer In-Reply-To: <35235.129.16.235.160.1062599298.squirrel@retposhto.ikso.net> Message-ID: On Wed, 3 Sep 2003 henning@ikso.net wrote: > Are anyone aware of a haskell html-colorizer a' la the emacs-mode? GNU Enscript (1.6 and newer) can produce html and not just postscript and it knows how to colorize haskell. -- /Dennis From GK@ninebynine.org Thu Sep 4 08:51:34 2003 From: GK@ninebynine.org (Graham Klyne) Date: Thu, 04 Sep 2003 08:51:34 +0100 Subject: Why not haskell? In-Reply-To: <200309040146.15981.nick.name@inwind.it> Message-ID: <5.1.0.14.2.20030904084802.026e5538@127.0.0.1> I would say that being a relatively niche technology would be a factor in not using Haskell, because finding staff to handle software maintenance is problematic. Also, I think that memory usage and performance is still an issue (real or imagined? My sense is that compiled Haskell can match widely languages like Perl or Python, but not yet Java or C/C++). #g -- At 01:46 04/09/03 +0200, Nick Name wrote: >Inspired by the various replies to the "haskell for non-haskell sake" I >wonder why so much people uses haskell "only for prototipying" or >producing code in other languages. > >I am just curios to hear from people who do not use haskell for project >releases, or just think it's not suitable for a mature project, what >exactly they find bad in current haskell implementations, or perhaps in >the standard. > >Vincenzo > >_______________________________________________ >Haskell-Cafe mailing list >Haskell-Cafe@haskell.org >http://www.haskell.org/mailman/listinfo/haskell-cafe ------------ Graham Klyne GK@NineByNine.org From bjpop@cs.mu.OZ.AU Fri Sep 5 09:14:33 2003 From: bjpop@cs.mu.OZ.AU (Bernard James POPE) Date: Fri, 5 Sep 2003 18:14:33 +1000 (EST) Subject: haskell reify, was sequencing data structures In-Reply-To: <3429668D0E777A499EE74A7952C382D1BE61B2@EUR-MSG-01.europe.corp.microsoft.com> from Simon Marlow at "Aug 26, 2003 04:05:27 pm" Message-ID: <200309050814.SAA21851@mulga.cs.mu.OZ.AU> > > However, the new debugger > > in GHC must do something to get names of things, so perhaps the > > -prof hack is no longer needed? > > I believe the debugger that Robert Ennals is working on tries to infer > constructor names from the symbol table of the binary. Personally, I'd > like to see debugging info placed in a separate segment of the binary, > so it won't be loaded with the code but can be used by a debugger. I wonder whether keeping the names of data constructors in info tables is expensive? I would have guessed not. (more debugging info such as function names and src locations is likely to be more expensive however). Anyway, would it be possible/feasible for GHC to keep names in info tables all the time, not just for profiling? At least for data-constructors? Cheers, Bernie. From simonmar@microsoft.com Fri Sep 5 10:39:47 2003 From: simonmar@microsoft.com (Simon Marlow) Date: Fri, 5 Sep 2003 10:39:47 +0100 Subject: haskell reify, was sequencing data structures Message-ID: <3429668D0E777A499EE74A7952C382D1C623E0@EUR-MSG-01.europe.corp.microsoft.com> =20 > I wonder whether keeping the names of data constructors in info tables > is expensive? >=20 > I would have guessed not. (more debugging info such as > function names and src locations is likely to be more=20 > expensive however). >=20 > Anyway, would it be possible/feasible for GHC to keep names=20 > in info tables > all the time, not just for profiling? At least for data-constructors? That's a distinct possibility. If you send patches we'll incorporate them. Cheers, Simon From hinsen@cnrs-orleans.fr Fri Sep 5 15:35:01 2003 From: hinsen@cnrs-orleans.fr (Konrad Hinsen) Date: Fri, 5 Sep 2003 16:35:01 +0200 Subject: Library documentation Message-ID: <200309051635.01019.hinsen@cnrs-orleans.fr> Is there any place where I can find an introductory documentation for the= =20 Haskell libraries? To be more concrete, say I am interested in using=20 Data.Array.ST.STUArray. I would need some explanation of the ST monad, an= d=20 then an explanation of how I can use an STUArray in an ST monad. Plus ide= ally=20 examples. But the same applies to most of the libraries. I can't find any= =20 documentation other than the very terse reference or the parts that happe= n to=20 be covered by textbooks. Konrad. From per@L4i.se Fri Sep 5 16:17:59 2003 From: per@L4i.se (Per Larsson) Date: Fri, 5 Sep 2003 17:17:59 +0200 Subject: Library documentation In-Reply-To: <200309051635.01019.hinsen@cnrs-orleans.fr> References: <200309051635.01019.hinsen@cnrs-orleans.fr> Message-ID: <200309051717.59950.per@L4i.se> On Friday 05 September 2003 16.35, Konrad Hinsen wrote: > Is there any place where I can find an introductory documentation for t= he > Haskell libraries? To be more concrete, say I am interested in using > Data.Array.ST.STUArray. I would need some explanation of the ST monad, = and > then an explanation of how I can use an STUArray in an ST monad. Plus > ideally examples. But the same applies to most of the libraries. I can'= t > find any documentation other than the very terse reference or the parts > that happen to be covered by textbooks. > > Konrad. I have the same problem and as far as I know there is only haddock genera= ted=20 html-documentation from sparse source comments available for many parts o= f=20 the library. But for some of the modules you can get exactly the informa= tion=20 you are asking for, e.g. the Parsec module has its own readable documenta= tion=20 at http://www.cs.uu.nl/~daan/parsec.html and the Control.Monad module has recently been superbly documented by John Newbern at=20 http://www.nomaware.com/monads/html/. Best Regards Per From GK@ninebynine.org Fri Sep 5 18:29:49 2003 From: GK@ninebynine.org (Graham Klyne) Date: Fri, 05 Sep 2003 18:29:49 +0100 Subject: Library documentation In-Reply-To: <200309051635.01019.hinsen@cnrs-orleans.fr> Message-ID: <5.1.0.14.2.20030905182622.00b82f08@127.0.0.1> I think the GHC documentation is a place to look, though it is a bit sparse in places. viz: http://www.haskell.org/ghc/docs/latest/html/base/index.html Hmmm, it sounds as if you already found that. I mention it because I didn't until someone pointed me at it. #g -- At 16:35 05/09/03 +0200, Konrad Hinsen wrote: >Is there any place where I can find an introductory documentation for the >Haskell libraries? To be more concrete, say I am interested in using >Data.Array.ST.STUArray. I would need some explanation of the ST monad, and >then an explanation of how I can use an STUArray in an ST monad. Plus ideally >examples. But the same applies to most of the libraries. I can't find any >documentation other than the very terse reference or the parts that happen to >be covered by textbooks. > >Konrad. > >_______________________________________________ >Haskell-Cafe mailing list >Haskell-Cafe@haskell.org >http://www.haskell.org/mailman/listinfo/haskell-cafe ------------ Graham Klyne GK@NineByNine.org From mark@chaos.x-philes.com Tue Sep 9 16:21:06 2003 From: mark@chaos.x-philes.com (Mark Carroll) Date: Tue, 9 Sep 2003 11:21:06 -0400 (EDT) Subject: Why not haskell? In-Reply-To: <200309040146.15981.nick.name@inwind.it> Message-ID: On Thu, 4 Sep 2003, Nick Name wrote: > I am just curios to hear from people who do not use haskell for project > releases, or just think it's not suitable for a mature project, what > exactly they find bad in current haskell implementations, or perhaps in > the standard. We're not really there yet, but we're likely to have to code stuff that works fast, guaranteed, with low overhead - for instance, in real-time embedded systems for military sensor data processing. I don't know where Haskell currently is, but I'm far from certain that it's up to that task. -- Mark From cmiltonperl@yahoo.com Tue Sep 9 16:54:57 2003 From: cmiltonperl@yahoo.com (Christopher Milton) Date: Tue, 9 Sep 2003 08:54:57 -0700 (PDT) Subject: Why not haskell? In-Reply-To: Message-ID: <20030909155457.60611.qmail@web20810.mail.yahoo.com> --- Mark Carroll wrote: > On Thu, 4 Sep 2003, Nick Name wrote: > > > I am just curios to hear from people who do not use haskell for project > > releases, or just think it's not suitable for a mature project, what > > exactly they find bad in current haskell implementations, or perhaps in > > the standard. > > We're not really there yet, but we're likely to have to code stuff that > works fast, guaranteed, with low overhead - for instance, in real-time > embedded systems for military sensor data processing. I don't know where > Haskell currently is, but I'm far from certain that it's up to that task. What about Embedded Gofer? ;-) "Lambdas in the Liftshaft" by Malcolm Wallace. Then there's the DARPA-funded Project Timber at OGI: http://www.cse.ogi.edu/PacSoft/projects/Timber/Default.htm Chris Milton From qrczak@knm.org.pl Tue Sep 9 19:56:48 2003 From: qrczak@knm.org.pl (Marcin 'Qrczak' Kowalczyk) Date: Tue, 9 Sep 2003 20:56:48 +0200 Subject: Prefix and postfix operators In-Reply-To: <9ED672B9D1A64C489291BE0FB822217D01FFAD4A@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> References: <9ED672B9D1A64C489291BE0FB822217D01FFAD4A@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Message-ID: <200309092056.48949.qrczak@knm.org.pl> Dnia ¶ro 27. sierpnia 2003 20:48, Eugene Nonko napisa³: > Are there any plans to extend Haskell parser to support prefix and > postfix operators? It will be just great for domain-specific languages. > It always kind of bothered me that unary minus is special. Having only infix operators has the advantage that operators are never adjacent to other operators, so you don't need extra spaces just to separate them nor complicated lexical conventions which tell where multi-character operator sequences should be split. Especially as prefix and postfix operators look better without spaces. In Haskell there are rare troubles because of - and ~ (lazy patterns). ! (strictness annotations) happens not to be used adjacent to operators. I think there are no more unary operators, so it's almost safe; problematic cases with - and ~ are really rare. -- __("< Marcin Kowalczyk \__/ qrczak@knm.org.pl ^^ http://qrnik.knm.org.pl/~qrczak/ From listener@thaldyron.com Wed Sep 10 11:33:50 2003 From: listener@thaldyron.com (Peter Robinson) Date: Wed, 10 Sep 2003 12:33:50 +0200 Subject: Standalone Parser for Haskell Grammar Message-ID: <200309101233.51274.listener@thaldyron.com> Hello! Does anyone know a reasonable standalone Parser for the Haskell Grammar? The only one i found was hsparser but it's still an alpha release and i get a few errors during compiling. I know i could write one using Happy but i don't want to reinvent the wheel... regards Peter From ross@soi.city.ac.uk Wed Sep 10 12:39:50 2003 From: ross@soi.city.ac.uk (Ross Paterson) Date: Wed, 10 Sep 2003 12:39:50 +0100 Subject: Standalone Parser for Haskell Grammar In-Reply-To: <200309101233.51274.listener@thaldyron.com> References: <200309101233.51274.listener@thaldyron.com> Message-ID: <20030910113950.GA5677@soi.city.ac.uk> On Wed, Sep 10, 2003 at 12:33:50PM +0200, Peter Robinson wrote: > Does anyone know a reasonable standalone Parser for the Haskell Grammar? The > only one i found was hsparser but it's still an alpha release and i get a few > errors during compiling. I know i could write one using Happy but i don't > want to reinvent the wheel... The latest version of hsparser is the Language.Haskell.* modules in the haskell-src package, included with recent releases of GHC and Hugs. From listener@thaldyron.com Wed Sep 10 15:45:59 2003 From: listener@thaldyron.com (Peter Robinson) Date: Wed, 10 Sep 2003 16:45:59 +0200 Subject: Standalone Parser for Haskell Grammar Message-ID: <200309101646.00702.listener@thaldyron.com> Thanks to both of you, you helped me a lot! :-) Peter On Wednesday 10 September 2003 14:56, you wrote: > You may find an error correcting combinator based parser in the > directory uust, to be found at cvs.cs.uu.nl > > Just check out the whole tree and follow the building and installing > instructions. The parser etc can be found in the UHC section. It makes > use of the parser combinators found elsewhere in the tree, > > Doaitse Swierstra > > On woensdag, september 10, 2003, at 12:33 PM, Peter Robinson wrote: > > Hello! > > Does anyone know a reasonable standalone Parser for the Haskell > > Grammar? The > > only one i found was hsparser but it's still an alpha release and i > > get a few > > errors during compiling. I know i could write one using Happy but i > > don't > > want to reinvent the wheel... > > regards > > Peter > > > > _______________________________________________ > > Haskell-Cafe mailing list > > Haskell-Cafe@haskell.org > > http://www.haskell.org/mailman/listinfo/haskell-cafe From doaitse@cs.uu.nl Wed Sep 10 13:56:14 2003 From: doaitse@cs.uu.nl (Doaitse Swierstra) Date: Wed, 10 Sep 2003 14:56:14 +0200 Subject: Standalone Parser for Haskell Grammar In-Reply-To: <200309101233.51274.listener@thaldyron.com> Message-ID: <28D30BE0-E38E-11D7-85DB-000A95729144@cs.uu.nl> You may find an error correcting combinator based parser in the directory uust, to be found at cvs.cs.uu.nl Just check out the whole tree and follow the building and installing instructions. The parser etc can be found in the UHC section. It makes use of the parser combinators found elsewhere in the tree, Doaitse Swierstra On woensdag, september 10, 2003, at 12:33 PM, Peter Robinson wrote: > Hello! > Does anyone know a reasonable standalone Parser for the Haskell > Grammar? The > only one i found was hsparser but it's still an alpha release and i > get a few > errors during compiling. I know i could write one using Happy but i > don't > want to reinvent the wheel... > regards > Peter > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe From kuq32tr02 at sneakemail.com Fri Sep 12 23:08:30 2003 From: kuq32tr02 at sneakemail.com (kuq32tr02@sneakemail.com) Date: Fri Sep 12 18:08:33 2003 Subject: Simple IO Monad problem Message-ID: <22197-22642@sneakemail.com> Hello, I'm starting to use Haskell for writing actual programs using monads and I'm already lost. I have the following script: #!/usr/bin/runhugs > module Main where > import System(getArgs) > main = do putStr "Hello, World\n" > strs <- getArgs > map putStrLn strs Which gives the following error: runhugs: Error occurred Reading file "./mailalias.lhs": Reading file "/usr/lib/hugs/lib/System.hs": Reading file "./mailalias.lhs": Type checking ERROR "./mailalias.lhs":5 - Type error in final generator *** Term : map putStrLn strs *** Type : [IO ()] *** Does not match : IO a Can someone please explain what I'm doing wrong? Thanks! From nick.name at inwind.it Sat Sep 13 01:25:39 2003 From: nick.name at inwind.it (Nick Name) Date: Fri Sep 12 18:21:36 2003 Subject: Simple IO Monad problem In-Reply-To: <22197-22642@sneakemail.com> References: <22197-22642@sneakemail.com> Message-ID: <200309130025.39933.nick.name@inwind.it> Alle 00:08, sabato 13 settembre 2003, kuq32tr02@sneakemail.com ha scritto: > ERROR "./mailalias.lhs":5 - Type error in final generator > *** Term ? ? ? ? ? : map putStrLn strs > *** Type ? ? ? ? ? : [IO ()] > *** Does not match : IO a > > Can someone please explain what I'm doing wrong? This is a type error. You have to "mapM_", not map, see the types in the documentation and try to convince yourself that you need an IO list and not a list of IO, then that you really need an IO () and not an IO list (wich would be returned by mapM). The syntax for a list of t in haskell is [t], in case you didn't notice. V. From angagon at earthlink.net Fri Sep 12 17:27:59 2003 From: angagon at earthlink.net (Matt O'Connor) Date: Fri Sep 12 18:24:37 2003 Subject: An IO Question from a Newbie Message-ID: <200309121627.59788.angagon@earthlink.net> Hello all. I'm new to functional programming and Haskell, but have been programming in C and Java for a while. I've been going through the tutorials and whatnot on haskell.org. I've read from the Gentle Introduction to Haskell about IO and some of the other stuff and I have a question about it. main = do putStr "Type Something: " str <- getLine putStrLn ("You typed: " ++ str) When compile and run this code the "Type Something: " isn't displayed until the putStrLn. So there is no prompt. The output looks like this. s Type Something: You typed: s But if I change the putStr "Type Something: " to a putStrLn or put a \n at the end of the string it displays the text immediately (ie, when I want it to). Is there a good reason for this? Am I doing something wrong? Or do I need to call some kind of standard output flush? Thanks. Oh, I'm using ghc. Matt -- "Not everone knows Josiah Royce's definition of a liar as a man who willfully misplaces his ontological predicates, but everyone who has ever told a lie will recognize its accuracy." Ch. 94: Truth - "The Great Ideas: A Synopticon of Great Books of the Western World" From ddarius at hotpop.com Fri Sep 12 19:25:07 2003 From: ddarius at hotpop.com (Derek Elkins) Date: Fri Sep 12 18:27:34 2003 Subject: Simple IO Monad problem In-Reply-To: <22197-22642@sneakemail.com> References: <22197-22642@sneakemail.com> Message-ID: <20030912182507.00001455.ddarius@hotpop.com> On 12 Sep 2003 22:08:30 -0000 kuq32tr02@sneakemail.com wrote: > Hello, > > I'm starting to use Haskell for writing actual programs using monads > and I'm already lost. > > I have the following script: > > #!/usr/bin/runhugs > > > module Main where > > import System(getArgs) > > main = do putStr "Hello, World\n" > > strs <- getArgs > > map putStrLn strs > > > Which gives the following error: > > runhugs: Error occurred > Reading file "./mailalias.lhs": > Reading file "/usr/lib/hugs/lib/System.hs": > Reading file "./mailalias.lhs": > Type checking > ERROR "./mailalias.lhs":5 - Type error in final generator > *** Term : map putStrLn strs > *** Type : [IO ()] > *** Does not match : IO a > > Can someone please explain what I'm doing wrong? > map :: (a -> b) -> [a] -> [b] putStrLn :: String -> IO () therefore map putStrLn :: [String] -> [IO ()] map maps a -pure function- over a list. What you want is to map a -monadic computation- over the list, further you don't care about the result. mapM_ :: Monad m => (a -> m b) -> [a] -> m () mapM_ putStrLn :: [String] -> IO () From hdaume at ISI.EDU Fri Sep 12 17:18:22 2003 From: hdaume at ISI.EDU (Hal Daume III) Date: Fri Sep 12 19:18:41 2003 Subject: An IO Question from a Newbie In-Reply-To: <200309121627.59788.angagon@earthlink.net> Message-ID: This is a buffering problem. Use hSetBuffering to fix this (see Chapter 3 in YAHT -- www.isi.edu/~hdaume/htut/). Alternatively, use: > main = do putStrLn "Type Something:" > ... in which case the "Ln" part will force it to be printed. - Hal On Fri, 12 Sep 2003, Matt O'Connor wrote: > Hello all. I'm new to functional programming and Haskell, but have been > programming in C and Java for a while. I've been going through the tutorials > and whatnot on haskell.org. I've read from the Gentle Introduction to > Haskell about IO and some of the other stuff and I have a question about it. > > main = do putStr "Type Something: " > str <- getLine > putStrLn ("You typed: " ++ str) > > When compile and run this code the "Type Something: " isn't displayed until > the putStrLn. So there is no prompt. The output looks like this. > > s > Type Something: You typed: s > > But if I change the putStr "Type Something: " to a putStrLn or put a \n at the > end of the string it displays the text immediately (ie, when I want it to). > Is there a good reason for this? Am I doing something wrong? Or do I need > to call some kind of standard output flush? Thanks. > > Oh, I'm using ghc. > > > Matt > > -- Hal Daume III | hdaume@isi.edu "Arrest this man, he talks in maths." | www.isi.edu/~hdaume From ddarius at hotpop.com Fri Sep 12 20:17:05 2003 From: ddarius at hotpop.com (Derek Elkins) Date: Fri Sep 12 19:19:17 2003 Subject: An IO Question from a Newbie In-Reply-To: <200309121627.59788.angagon@earthlink.net> References: <200309121627.59788.angagon@earthlink.net> Message-ID: <20030912191705.0000677a.ddarius@hotpop.com> On Fri, 12 Sep 2003 16:27:59 -0600 Matt O'Connor wrote: > Hello all. I'm new to functional programming and Haskell, but have > been programming in C and Java for a while. I've been going through > the tutorials and whatnot on haskell.org. I've read from the Gentle > Introduction to Haskell about IO and some of the other stuff and I > have a question about it. > > main = do putStr "Type Something: " > str <- getLine > putStrLn ("You typed: " ++ str) > > When compile and run this code the "Type Something: " isn't displayed > until the putStrLn. So there is no prompt. The output looks like > this. > > s > Type Something: You typed: s > > But if I change the putStr "Type Something: " to a putStrLn or put a > \n at the end of the string it displays the text immediately (ie, when > I want it to). Is there a good reason for this? Am I doing something > wrong? Or do I need to call some kind of standard output flush? Yes, the problem is that the first line is not being flushed. Importing hFlush and stdout from IO (or System.IO) and adding hFlush stdout after you putStr will fix it. Alternatively, you could change the buffering on stdout to NoBuffering. I don't know if there is a standard putStrWithFlush function or just a better answer than this. A putStrWithFlush would be trivial to write or more generally to a HOF that flushed after performing the action passed to it. From glynn.clements at virgin.net Sat Sep 13 01:42:41 2003 From: glynn.clements at virgin.net (Glynn Clements) Date: Fri Sep 12 19:43:04 2003 Subject: An IO Question from a Newbie In-Reply-To: <200309121627.59788.angagon@earthlink.net> References: <200309121627.59788.angagon@earthlink.net> Message-ID: <16226.23025.91228.835965@cerise.nosuchdomain.co.uk> Matt O'Connor wrote: > Hello all. I'm new to functional programming and Haskell, but have been > programming in C and Java for a while. I've been going through the tutorials > and whatnot on haskell.org. I've read from the Gentle Introduction to > Haskell about IO and some of the other stuff and I have a question about it. > > main = do putStr "Type Something: " > str <- getLine > putStrLn ("You typed: " ++ str) > > When compile and run this code the "Type Something: " isn't displayed until > the putStrLn. So there is no prompt. The output looks like this. > > s > Type Something: You typed: s > > But if I change the putStr "Type Something: " to a putStrLn or put a \n at the > end of the string it displays the text immediately (ie, when I want it to). > Is there a good reason for this? Am I doing something wrong? Or do I need > to call some kind of standard output flush? Thanks. Yes; if you don't want a newline after the prompt, you need to use: hFlush stdout to flush the stream. -- Glynn Clements From brandon at its.caltech.edu Sun Sep 14 01:23:39 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Sun Sep 14 03:23:46 2003 Subject: An IO Question from a Newbie In-Reply-To: References: Message-ID: Hal was pretty terse, so I'll explain why switching to putStrLn will help. stdout is line buffered. At least by default (see hSetBuffering). That means output will only be flushed to the screen once a newline is written. Your prompt wasn't printed because it didn't have a newline, so it was buffered until the second print provided one (read from the user, by way of s). This is hardly specific to Haskell. Try this C program: #import #import int main(int a,char** b) { fputs("test",stdout); sleep(1); fputs("boo!\n",stdout); return 0; } on my system this sleeps a second then prints "testboo!". I used fputs rather than puts becuase puts appends a newline and fputs doesn't. Similarly, putStrLn appends a newline and putStr doesn't. This change does mean input will be entered on the next line, rather then on the same line as the prompt. This should be acceptable. If you really need to have a prompt at the start of the line then consider hFlush or hSetBuffering. Every time I've seen this question asked people post responses that only mention hFlush, implying you need it to display output when you want it. Please don't give newbies the impression you need hFlush for basic IO. Monadic IO is scary enough to begin with. This seems to be a common problem so I added some comments about it on the Wiki, under UsingIO in the FrequentlyAskedQuestions. Brandon On Fri, 12 Sep 2003, Hal Daume III wrote: > This is a buffering problem. Use hSetBuffering to fix this (see Chapter 3 > in YAHT -- www.isi.edu/~hdaume/htut/). Alternatively, use: > > > main = do putStrLn "Type Something:" > > ... > > > in which case the "Ln" part will force it to be printed. > > - Hal > > On Fri, 12 Sep 2003, Matt O'Connor wrote: > > > Hello all. I'm new to functional programming and Haskell, but have been > > programming in C and Java for a while. I've been going through the tutorials > > and whatnot on haskell.org. I've read from the Gentle Introduction to > > Haskell about IO and some of the other stuff and I have a question about it. > > > > main = do putStr "Type Something: " > > str <- getLine > > putStrLn ("You typed: " ++ str) > > > > When compile and run this code the "Type Something: " isn't displayed until > > the putStrLn. So there is no prompt. The output looks like this. > > > > s > > Type Something: You typed: s > > > > But if I change the putStr "Type Something: " to a putStrLn or put a \n at the > > end of the string it displays the text immediately (ie, when I want it to). > > Is there a good reason for this? Am I doing something wrong? Or do I need > > to call some kind of standard output flush? Thanks. > > > > Oh, I'm using ghc. > > > > > > Matt > > > > > > -- > Hal Daume III | hdaume@isi.edu > "Arrest this man, he talks in maths." | www.isi.edu/~hdaume > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > > From qrczak at knm.org.pl Sun Sep 14 12:10:58 2003 From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk) Date: Sun Sep 14 05:11:01 2003 Subject: No safety in numbers In-Reply-To: <200308212230.33161.hinsen@cnrs-orleans.fr> References: <200308212230.33161.hinsen@cnrs-orleans.fr> Message-ID: <200309141110.58099.qrczak@knm.org.pl> Dnia czw 21. sierpnia 2003 22:30, Konrad Hinsen napisa?: > k_B = 0.0083144708636327096 > > The trouble is that k_B then becomes "Double" by default (or any other > type I declare it to be). Declare it as k_B :: Fractional a => a Performance may suffer even in GHC if optimization is turned off. BTW, Hugs incorrectly converts such numbers through Double which it represents as C float. -- __("< Marcin Kowalczyk \__/ qrczak@knm.org.pl ^^ http://qrnik.knm.org.pl/~qrczak/ From mark at felinemail.zzn.com Sun Sep 14 10:40:51 2003 From: mark at felinemail.zzn.com (Mark Espinoza) Date: Sun Sep 14 10:28:18 2003 Subject: YAHT Problem Message-ID: <33B50AEEEAD0E4040B5E8D635DE618A6@mark.felinemail.zzn.com> An HTML attachment was scrubbed... URL: http://porky.devel.redhat.com/pipermail/haskell-cafe/attachments/20030914/e79f9725/attachment.htm From glynn.clements at virgin.net Sun Sep 14 17:09:37 2003 From: glynn.clements at virgin.net (Glynn Clements) Date: Sun Sep 14 11:42:59 2003 Subject: An IO Question from a Newbie In-Reply-To: References: Message-ID: <16228.33969.14389.53704@cerise.nosuchdomain.co.uk> Brandon Michael Moore wrote: > Hal was pretty terse, so I'll explain why switching to putStrLn will help. > > stdout is line buffered. > > At least by default (see hSetBuffering). That means output will only be > flushed to the screen once a newline is written. Your prompt wasn't > printed because it didn't have a newline, so it was buffered until the > second print provided one (read from the user, by way of s). > > This is hardly specific to Haskell. Try this C program: But there's one significant difference between C and Haskell, which is applicable in the case of Matt's program. In C, any line-buffered output streams are automatically flushed when a read from an unbuffered or line-buffered stream can't be satisfied from its buffer. Also, it seemed fairly clear from Matt's original message that: a) he didn't want to have to force a new-line (he noted that doing so eliminated the problem), and b) he understood the concept of flushing, but presumably didn't know how to do it in Haskell. While we're on the subject, I'll point out a couple of other differences between the buffering in ANSI C's stdio library and Haskell's: 1. In Haskell, you can change the buffering mode at any point; in C, you have to change it before any data is read from or written to the stream, otherwise the behaviour is undefined. 2. For an input stream which is associated with a tty, changing the buffering mode may also change the terminal settings (setting it to unbuffered disables canonical mode while setting it to line-buffered or fully-buffered enables it). -- Glynn Clements From brandon at its.caltech.edu Sun Sep 14 15:11:15 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Sun Sep 14 17:11:19 2003 Subject: An IO Question from a Newbie In-Reply-To: <16228.33969.14389.53704@cerise.nosuchdomain.co.uk> References: <16228.33969.14389.53704@cerise.nosuchdomain.co.uk> Message-ID: On Sun, 14 Sep 2003, Glynn Clements wrote: > > Brandon Michael Moore wrote: > > > Hal was pretty terse, so I'll explain why switching to putStrLn will help. > > > > stdout is line buffered. > > > > At least by default (see hSetBuffering). That means output will only be > > flushed to the screen once a newline is written. Your prompt wasn't > > printed because it didn't have a newline, so it was buffered until the > > second print provided one (read from the user, by way of s). > > > > This is hardly specific to Haskell. Try this C program: > > But there's one significant difference between C and Haskell, which is > applicable in the case of Matt's program. In C, any line-buffered > output streams are automatically flushed when a read from an > unbuffered or line-buffered stream can't be satisfied from its buffer. Interesting. I didn't know this. Maybe we should match this behaviour, or provide a write-string-and-flush function. It seems like this issue is causing an undue amound of trouble. > Also, it seemed fairly clear from Matt's original message that: > > a) he didn't want to have to force a new-line (he noted that doing so > eliminated the problem), and I should note here that there is a gnu readline binding distributed with GHC. It's undocumented, but it seems to follow the C API pretty closely, and you can make a decent interface using only two of the functions. > b) he understood the concept of flushing, but presumably didn't know > how to do it in Haskell. > > While we're on the subject, I'll point out a couple of other > differences between the buffering in ANSI C's stdio library and > Haskell's: > > 1. In Haskell, you can change the buffering mode at any point; in C, > you have to change it before any data is read from or written to the > stream, otherwise the behaviour is undefined. > > 2. For an input stream which is associated with a tty, changing the > buffering mode may also change the terminal settings (setting it to > unbuffered disables canonical mode while setting it to line-buffered > or fully-buffered enables it). > > -- > Glynn Clements > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > > From droundy at abridgegame.org Sun Sep 14 19:52:59 2003 From: droundy at abridgegame.org (David Roundy) Date: Sun Sep 14 18:53:00 2003 Subject: zlib binding providing Handle to compressed file. Message-ID: <20030914225256.GA27346@jdj5.mit.edu> I've just created a binding to the gzopen function of zlib which causes it to create a Handle. The code currently only supports ReadMode and WriteMode, and hFileSize won't work properly when reading a file. In fact, pretty much nothing but plain old reading and writing will work, but such is life. Anyhow in case anyone is interested, I'm attaching the code. It creates a pipe using pipe(2) and spawns a thread to pass the data between the pipe and gzread or gzwrite. It's not pretty, but it's better than any other solution I could think of. Suggestions or criticisms are welcome. -- David Roundy http://www.abridgegame.org/darcs -------------- next part -------------- \begin{code} module Zlib ( gzOpenFile, gzWriteFile, gzReadFile ) where import IO import System.IO ( hGetBuf, hPutBuf ) import Control.Concurrent ( forkIO ) import Monad ( when ) import Foreign.C.String ( CString, withCString ) import Foreign.Marshal.Array ( mallocArray, withArray, peekArray ) import Foreign.Marshal.Alloc ( free ) import Foreign.Ptr ( Ptr ) import Data.Word import GHC.Handle ( openFd ) fdToReadHandle fd fn = openFd fd Nothing fn ReadMode False False fdToWriteHandle fd fn = openFd fd Nothing fn WriteMode False False gzOpenFile :: FilePath -> IOMode -> IO Handle gzWriteFile :: FilePath -> String -> IO () gzOpenFile f ReadMode = withCString f $ \fstr -> withCString "rb" $ \rb-> do gzf <- c_gzopen fstr rb withArray [0,0] $ \fds -> do err <- c_pipe fds when (err /= 0) $ error "Pipe problem!" [infd,outfd] <- peekArray 2 fds writeH <- fdToWriteHandle (fromIntegral outfd) f buf <- mallocArray 1024 forkIO $ gzreader gzf writeH buf fdToReadHandle (fromIntegral infd) f where gzreader gzf h buf = do done <- hIsClosed h if done then do c_gzclose gzf free buf hClose h else do l <- c_gzread gzf buf 1024 hPutBuf h buf l if l < 1024 then do free buf c_gzclose gzf hClose h else gzreader gzf h buf gzOpenFile f WriteMode = withCString f $ \fstr -> withCString "wb" $ \wb-> do gzf <- c_gzopen fstr wb withArray [0,0] $ \fds -> do err <- c_pipe fds when (err /= 0) $ error "Pipe problem!" [infd,outfd] <- peekArray 2 fds readH <- fdToReadHandle (fromIntegral infd) f buf <- mallocArray 1024 forkIO $ gzwriter gzf readH buf fdToWriteHandle (fromIntegral outfd) f where gzwriter gzf h buf = do done <- hIsEOF h if done then do c_gzclose gzf free buf hClose h else do l <- hGetBuf h buf 1024 c_gzwrite gzf buf l gzwriter gzf h buf gzWriteFile f s = do h <- gzOpenFile f WriteMode hPutStr h s hClose h gzReadFile f s = do h <- gzOpenFile f WriteMode hGetContents h foreign import ccall unsafe "static unistd.h pipe" c_pipe :: Ptr Int -> IO Int foreign import ccall unsafe "static unistd.h read" c_read :: Ptr Word8 -> Int -> IO Int foreign import ccall unsafe "static zlib.h gzopen" c_gzopen :: CString -> CString -> IO (Ptr ()) foreign import ccall unsafe "static zlib.h gzclose" c_gzclose :: Ptr () -> IO () foreign import ccall unsafe "static zlib.h gzread" c_gzread :: Ptr () -> Ptr Word8 -> Int -> IO Int foreign import ccall unsafe "static zlib.h gzwrite" c_gzwrite :: Ptr () -> Ptr Word8 -> Int -> IO () \end{code} From stein at eecs.harvard.edu Mon Sep 15 01:48:29 2003 From: stein at eecs.harvard.edu (Lex Stein) Date: Mon Sep 15 00:48:30 2003 Subject: threaded red-black tree Message-ID: Hi, No one responded to my question several weeks ago about a purely functional implementation of a threaded, red-black tree. My message was sent about the same time as that flurry of silly emails about "how to respond to homework questions". Was my message not responded to because it was classified as a homework question? I assure you this is officework, not homework. I ended up porting Okasaki's red-black tree implementation; hacking it apart with a bunch of mutation for the threading of the list through the tree. However, I'm still missing a deletion function and I haven't been able to find a prototype (Okasaki's red-black tree module lacks delete). My study of the RB-tree deletion routine in CLR hasn't yet yielded an adaptation for a functional environment. Any advice would be much appreciated. Thanks, Lex -- Lex Stein http://www.eecs.harvard.edu/~stein/ stein@eecs.harvard.edu cell: 617-233-0246 From cwitty at newtonlabs.com Sun Sep 14 23:44:56 2003 From: cwitty at newtonlabs.com (Carl Witty) Date: Mon Sep 15 01:53:29 2003 Subject: threaded red-black tree In-Reply-To: References: Message-ID: <1063604695.994.55.camel@flare> On Sun, 2003-09-14 at 21:48, Lex Stein wrote: > Hi, No one responded to my question several weeks ago about a purely > functional implementation of a threaded, red-black tree. I'm not sure what you mean by "threaded". By simply ignoring that word, I come up with the following solution :-) There is a purely functional implementation of a red-black tree in the MetaPRL system (www.metaprl.org), written in OCaml. For the latest CVS version of this red-black tree code, go to http://cvs.metaprl.org:12000/cvsweb/metaprl/libmojave/stdlib/ (look at lm_map.ml and lm_set.ml). Carl Witty From nick.name at inwind.it Mon Sep 15 15:38:38 2003 From: nick.name at inwind.it (Nick Name) Date: Mon Sep 15 08:34:42 2003 Subject: Instance checking and phantom types Message-ID: <200309151438.38891.nick.name@inwind.it> Hi all, I have an example wich I don't understand: ---------------begin class C t data T = T instance C T data C t => T1 t = T1 f1 :: T1 () f1 = T1 data C t => T2 t = T2 t f2 :: T2 () f2 = T2 () ----------------end The first function, f1, is accepted both by hugs and ghc, unlike the second wich is rejected. Why does this happen? Shouldn't f1 be rejected with "no instance C ()" Thanks for help Vincenzo From andrew at acooke.org Mon Sep 15 09:57:58 2003 From: andrew at acooke.org (andrew cooke) Date: Mon Sep 15 08:58:05 2003 Subject: threaded red-black tree In-Reply-To: References: Message-ID: <33731.139.229.3.218.1063630678.squirrel@127.0.0.1> this might help: http://www.cs.kent.ac.uk/people/staff/smk/redblack/Untyped.hs andrew Lex Stein said: > > Hi, No one responded to my question several weeks ago about a purely > functional implementation of a threaded, red-black tree. My message was > sent about the same time as that flurry of silly emails about "how to > respond to homework questions". Was my message not responded to because it > was classified as a homework question? I assure you this is officework, > not homework. I ended up porting Okasaki's red-black tree implementation; > hacking it apart with a bunch of mutation for the threading of the list > through the tree. However, I'm still missing a deletion function and I > haven't been able to find a prototype (Okasaki's red-black tree module > lacks delete). My study of the RB-tree deletion routine in CLR hasn't yet > yielded an adaptation for a functional environment. Any advice would be > much appreciated. > > Thanks, > Lex > > -- > Lex Stein http://www.eecs.harvard.edu/~stein/ > stein@eecs.harvard.edu cell: 617-233-0246 > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > > -- http://www.acooke.org/andrew From belloc_f at epita.fr Mon Sep 15 16:55:29 2003 From: belloc_f at epita.fr (Frederic BELLOC) Date: Mon Sep 15 09:55:35 2003 Subject: Read file problems Message-ID: HI, I have a recurrent probleme with function readFile and Monad : it is that i try to get a line, transform it and stock it in a list and what i want is that function return me the list. But hugs say me that in readFile my_file >>= \s -> map cons_line (lines s) readFile is a IO String type but is used here as [[Char]] type... and i don't know what to do... Ghc say me that is the lambda abstraction that fails with "map cons_line (lines s)" , he couldn't match IO against [] ! Help me please, i understand well (i hope) how Monad work but here i don't see a solution. Sorry for my poor english but i'm already a student. Thanks for all. -- C, socks and sun ;-) and haskell bugs :-( From hdaume at ISI.EDU Mon Sep 15 08:19:28 2003 From: hdaume at ISI.EDU (Hal Daume III) Date: Mon Sep 15 10:20:22 2003 Subject: Read file problems In-Reply-To: References: Message-ID: <1063635568.26187.16.camel@hal> Hi, This is a pretty simple problem to fix. (>>=) has type IO a -> (a -> IO b) -> IO b. 'readFile my_file' has type IO String, so this means whatever comes on the RHS of >>= should have type (String -> IO b). In your case, it doesn't. It has type String -> [something], but the [something] isn't an IO type. Hint: you need to put a call to 'return' in there. - Hal On Mon, 2003-09-15 at 06:55, Frederic BELLOC wrote: > HI, > I have a recurrent probleme with function readFile and Monad : > it is that i try to get a line, transform it and stock it in a list > and what i want is that function return me the list. > But hugs say me that in > > readFile my_file >>= \s -> map cons_line (lines s) > readFile is a IO String type but is used here as [[Char]] type... > and i don't know what to do... > > Ghc say me that is the lambda abstraction that fails with > "map cons_line (lines s)" , he couldn't match IO against [] ! > > Help me please, i understand well (i hope) how Monad work but here i don't see > a solution. > > Sorry for my poor english but i'm already a student. > > Thanks for all. -- -- Hal Daume III | hdaume@isi.edu "Arrest this man, he talks in maths." | www.isi.edu/~hdaume From belloc_f at epita.fr Mon Sep 15 17:27:58 2003 From: belloc_f at epita.fr (Frederic BELLOC) Date: Mon Sep 15 10:28:02 2003 Subject: Read file problems In-Reply-To: <1063635568.26187.16.camel@hal> (Hal Daume, III's message of "15 Sep 2003 07:19:28 -0700") References: <1063635568.26187.16.camel@hal> Message-ID: Hal Daume III disait r?cemment : > Hi, > > This is a pretty simple problem to fix. (>>=) has type IO a -> (a -> IO > b) -> IO b. 'readFile my_file' has type IO String, so this means > whatever comes on the RHS of >>= should have type (String -> IO b). In > your case, it doesn't. It has type String -> [something], but the > [something] isn't an IO type. > > Hint: you need to put a call to 'return' in there. > Yes, thanks ps : i've read something like that in the Monad Tutorial but i was not sure... http://www.nomaware.com/monads/html/index.html -- C, socks and sun ;-) No haskell bug :-D From simonmar at microsoft.com Mon Sep 15 17:11:56 2003 From: simonmar at microsoft.com (Simon Marlow) Date: Mon Sep 15 11:12:07 2003 Subject: An IO Question from a Newbie Message-ID: <3429668D0E777A499EE74A7952C382D1CF693D@EUR-MSG-01.europe.corp.microsoft.com> > > But there's one significant difference between C and > Haskell, which is > > applicable in the case of Matt's program. In C, any line-buffered > > output streams are automatically flushed when a read from an > > unbuffered or line-buffered stream can't be satisfied from > its buffer. > > Interesting. I didn't know this. Maybe we should match this > behaviour, or > provide a write-string-and-flush function. It seems like this issue > is causing an undue amound of trouble. I wrote GHC's IO library, and deliberately didn't include this feature. The previous version of the library did have such a feature, specifically for stdin/stdout. Note that the report doesn't say we must do this. The reason I didn't include the feature is because I can't see a way to do it right. Flushing *all* line-buffered handles (the ANSI C way) doesn't seem right. Flushing stdout just because we read from stdin is not right, because the two streams might refer to completely different I/O objects. Perhaps we should attempt to detect when there are two streams connected to the same I/O object (file, pipe, tty, whatever) and enable the magic flushing then. But now do I have to explain to people how this works? I suppose we could take the view that extra flushing is basically harmless, so it doesn't matter that we flush a bunch of Handles more often than we need to. The advantage of the current scheme is that it is easy to understand and explain; the library doesn't try to do clever stuff behind your back. The disadvantage is that it catches people out, and it sometimes requires you to import IO in an otherwise Prelude-only program. I'm more-or-less agnostic - if there's a way to avoid catching people out without introducing too much overhead or complicated rules that we have to explain, then I'll happily implement it. Cheers, Simon From Sven.Panne at informatik.uni-muenchen.de Mon Sep 15 21:07:45 2003 From: Sven.Panne at informatik.uni-muenchen.de (Sven Panne) Date: Mon Sep 15 14:08:28 2003 Subject: Instance checking and phantom types In-Reply-To: <200309151438.38891.nick.name@inwind.it> References: <200309151438.38891.nick.name@inwind.it> Message-ID: <3F65FFF1.7000101@informatik.uni-muenchen.de> Nick Name wrote: > Hi all, I have an example wich I don't understand: First of all, let's rename the constructors and types a bit to make things clearer add the instance in question, and remove the type signatures: ---------------------------------------------------------------- module Main where class C t data T = MkT instance C T instance C () data C t => T1 t = MkT1 f1 = MkT1 data C t => T2 t = MkT2 t f2 = MkT2 () ---------------------------------------------------------------- Then we can easily ask GHC: ---------------------------------------------------------------- panne@jeanluc:~> ghci -v0 Main.hs *Main> :i T1 MkT1 f1 T2 MkT2 f2 -- T1 is a type constructor, defined at Main.hs:8 data (C t) => T1 t = MkT1 -- MkT1 is a data constructor, defined at Main.hs:8 MkT1 :: forall t. T1 t -- f1 is a variable, defined at Main.hs:10 f1 :: forall t. T1 t -- T2 is a type constructor, defined at Main.hs:12 data (C t) => T2 t = MkT2 t -- MkT2 is a data constructor, defined at Main.hs:12 MkT2 :: forall t. (C t) => t -> T2 t -- f2 is a variable, defined at Main.hs:14 f2 :: T2 () ---------------------------------------------------------------- > The first function, f1, is accepted both by hugs and ghc, unlike the > second wich is rejected. > > Why does this happen? Shouldn't f1 be rejected with "no instance C ()" The reason is buried in http://haskell.org/onlinereport/decls.html#sect4.2.1 In a nutshell: The context in datatype declarations has only an effect for the *data* constructors of that type which use the type variables mentioned in the context. Contexts have no effect for the *type* constructor. IIRC the reason for this design decision was that contexts in type signatures should always be explicit. Cheers, S. From nick.name at inwind.it Mon Sep 15 21:44:21 2003 From: nick.name at inwind.it (Nick Name) Date: Mon Sep 15 14:40:14 2003 Subject: Instance checking and phantom types In-Reply-To: <3F65FFF1.7000101@informatik.uni-muenchen.de> References: <200309151438.38891.nick.name@inwind.it> <3F65FFF1.7000101@informatik.uni-muenchen.de> Message-ID: <200309152044.21533.nick.name@inwind.it> Alle 20:07, luned? 15 settembre 2003, Sven Panne ha scritto: > IIRC the > reason for this design decision was that contexts in type signatures > should always be explicit. Got it ;) Thanks for prompt reply. What does "should always be explicit" mean? Is there a notion of "explicit context" that I should know? Thanks Vincenzo From heringto at cs.unc.edu Mon Sep 15 15:50:13 2003 From: heringto at cs.unc.edu (Dean Herington) Date: Mon Sep 15 14:50:16 2003 Subject: simpler I/O buffering [was: RE: An IO Question from a Newbie] In-Reply-To: <3429668D0E777A499EE74A7952C382D1CF693D@EUR-MSG-01.europe.corp.microsoft.com> Message-ID: I've long thought that I/O buffering behavior--not just in Haskell, but in most places I've encountered it--was unnecessarily complicated. Perhaps it could be simplified dramatically by treating it as strictly a performance optimization. Here's a sketch of the approach. Writing a sequence of characters across the interface I'm proposing is a request by the writing program for those characters to appear at their destination "soon". Ideally, "soon" would be "immediately"; however, the characters' appearance may deliberately be delayed ("buffered"), for efficiency, as long as such delay is "unobtrusive" to a human user of the program. Buffering timeouts would depend on the device; for a terminal, perhaps 50-100 ms would be appropriate. Such an interval would tend not to be noticeable to a human user but would be long enough to effectively collect, say, an entire line of output for output "in one piece". The use of a reasonable timeout would avoid the confusing behavior where a newline-less prompt doesn't appear until the prompted data is entered. With this scheme, I/O buffering no longer has any real semantic content. (In particular, the interface never guarantees indefinite delay in outputting written characters. Line buffering, if semantically important, needs to be done above the level of this interface.) Hence, buffering control could be completely eliminated from the interface. However, I would retain it to provide (non-semantic) control over buffering. The optional buffer size currently has such an effect. A timeout value could be added for fine tuning. (Note that a timeout of zero would have an effect similar to Haskell's current NoBuffering.) Lastly, the "flush" operation would remain, as a hint that it's not worth waiting even the limited timeout period before endeavoring to make the characters appear. Is such an approach feasible? Has it been implemented anywhere? Would such behavior best be implemented by the operating system? Could it be implemented by the runtime system? Dean From Sven.Panne at informatik.uni-muenchen.de Mon Sep 15 22:04:41 2003 From: Sven.Panne at informatik.uni-muenchen.de (Sven Panne) Date: Mon Sep 15 15:05:25 2003 Subject: Instance checking and phantom types In-Reply-To: <200309152044.21533.nick.name@inwind.it> References: <200309151438.38891.nick.name@inwind.it> <3F65FFF1.7000101@informatik.uni-muenchen.de> <200309152044.21533.nick.name@inwind.it> Message-ID: <3F660D49.30104@informatik.uni-muenchen.de> Nick Name wrote: > Got it ;) Thanks for prompt reply. What does "should always be explicit" > mean? Is there a notion of "explicit context" that I should know? What I meant was the fact that you always have to write down *all* contexts involved in a type signature. Nothing is "inherited under the hood" by contexts in datatype declarations. Cheers, S. From Tom.Pledger at peace.com Tue Sep 16 09:36:52 2003 From: Tom.Pledger at peace.com (Tom Pledger) Date: Mon Sep 15 16:37:06 2003 Subject: threaded red-black tree In-Reply-To: <1063604695.994.55.camel@flare> References: <1063604695.994.55.camel@flare> Message-ID: <16230.8932.253663.5756@tux-17.corp.peace.com> Carl Witty writes: | On Sun, 2003-09-14 at 21:48, Lex Stein wrote: | > Hi, No one responded to my question several weeks ago about a purely | > functional implementation of a threaded, red-black tree. | | I'm not sure what you mean by "threaded". By simply ignoring that word, | I come up with the following solution :-) IIRC a threaded tree (in an imperative language) is one where 'unused' child pointers are made to point to the next (right child pointer) or previous (left child pointer) node in an in-order traversal. The pointers have to be flagged to show whether they're normal or threading. For example, if this tree R / \ L S \ M were threaded, the unused pointers would be pressed into service thus: L's left: still nil, because L is the least element M's left: thread to L (predecessor) M's right: thread to R (successor) S's left: thread to R (predecessor) S's right: still nil, because S is the greatest element A benefit of threading is that you can make an in-order traversal in constant space, because you don't have to remember the whole path from the root to your current position. You *could* translate it into a purely functional representation, along these lines data TRBT a = Empty | Node Colour (Ptr a) a (Ptr a) data Colour = Red | Black data Ptr a = Child (TRBT a) | Thread (TRBT a) but you'd have to do some tricky stuff http://haskell.org/hawiki/TyingTheKnot to deal with the circular nature of the data structure (e.g. in the above tree, R points to L, L points to M, and M points to R). Also note that every insert or delete is O(n) instead of O(log n), because *every* node in the old tree can see the part you change. I suggest you (Lex) either go imperative (with STRef or IORef) or do without threading, unless you're sure that you'll be doing many more lookups and traversals than inserts and deletes. Regards, Tom From glynn.clements at virgin.net Tue Sep 16 03:04:01 2003 From: glynn.clements at virgin.net (Glynn Clements) Date: Mon Sep 15 21:04:27 2003 Subject: simpler I/O buffering [was: RE: An IO Question from a Newbie] In-Reply-To: References: <3429668D0E777A499EE74A7952C382D1CF693D@EUR-MSG-01.europe.corp.microsoft.com> Message-ID: <16230.24961.810923.837752@cerise.nosuchdomain.co.uk> Dean Herington wrote: > I've long thought that I/O buffering behavior--not just in Haskell, but > in most places I've encountered it--was unnecessarily complicated. > Perhaps it could be simplified dramatically by treating it as strictly a > performance optimization. This isn't entirely possible; there will always be situations where it matters as to exactly when and how the data gets passed to the OS. My experience taught me that the simplest solution was never to use ANSI stdio buffering in such situations. > Here's a sketch of the approach. > > Writing a sequence of characters across the interface I'm proposing is a > request by the writing program for those characters to appear at their > destination "soon". Ideally, "soon" would be "immediately"; however, the > characters' appearance may deliberately be delayed ("buffered"), for > efficiency, as long as such delay is "unobtrusive" to a human user of the > program. Buffering timeouts would depend on the device; for a terminal, > perhaps 50-100 ms would be appropriate. Such an interval would tend not > to be noticeable to a human user but would be long enough to effectively > collect, say, an entire line of output for output "in one piece". The use > of a reasonable timeout would avoid the confusing behavior where a > newline-less prompt doesn't appear until the prompted data is entered. > > With this scheme, I/O buffering no longer has any real semantic content. > (In particular, the interface never guarantees indefinite delay in > outputting written characters. Line buffering, if semantically important, > needs to be done above the level of this interface.) That's already true, at least in C: if you output a line which is longer than the buffer, the buffer will be flushed before it contains a newline (i.e. the line won't be written atomically). > Hence, buffering > control could be completely eliminated from the interface. However, I > would retain it to provide (non-semantic) control over buffering. The > optional buffer size currently has such an effect. A timeout value could > be added for fine tuning. (Note that a timeout of zero would have an > effect similar to Haskell's current NoBuffering.) Lastly, the "flush" > operation would remain, as a hint that it's not worth waiting even the > limited timeout period before endeavoring to make the characters appear. > > Is such an approach feasible? Possibly. As things stand, anyone who writes code which relies upon output being held back until a flush is asking for trouble. So, your approach wouldn't make it any harder to write correct code, although it might make it significantly more obvious if code was incorrect. AFAICT, the biggest problem would be providing an upper bound on the delay, as that implies some form of preemptive concurrency. > Has it been implemented anywhere? Not that I know of. > Would such behavior best be implemented by the operating system? No. The OS (i.e. kernel) doesn't know anything about user-space buffering. Furthermore, one of the main functions of user-space buffering is to minimise the number of system calls, so putting it into the OS would be pointless. > Could it be implemented by the runtime system? It depends what you mean by "the runtime system"; it would have to be implemented in user-space. -- Glynn Clements From simons at cryp.to Tue Sep 16 14:37:48 2003 From: simons at cryp.to (Peter Simons) Date: Tue Sep 16 07:37:51 2003 Subject: Fast I/O with sockets Message-ID: <87oexl6k1f.fsf@peti.cryp.to> Hi, I was wondering how to read and write from a Socket in the most performant way. According to the library documentation from GHC, 'hGetBuf' and 'hPutBuf' are the way to go. But my problem with these functions is that they expect to read exactly the number of bytes I gave them! The IO computation hGetBuf sock ptr 1024 wants to read 1024 bytes, no less. This doesn't work for me, because I need to drive an interactive line-based protocol session. I need the read operation to consume as much data as is available and then to return immediately, so that I can process the command right away. As an alternative, I looked at 'hGetLine', but this doesn't work for me either, at least not very well, because in my protocol, lines are terminated by "\r\n" -- not just "\n", like hGetLine expects it. Finally, I considered 'hGetContents', which has the nice property of returning a lazy-evaluated list of the data read from the socket. But unfortunately, I don't see how I could implement something like a "read timeout" on the socket when using this function. Any ideas, anybody? Peter From simonmar at microsoft.com Tue Sep 16 14:45:47 2003 From: simonmar at microsoft.com (Simon Marlow) Date: Tue Sep 16 08:45:52 2003 Subject: Fast I/O with sockets Message-ID: <3429668D0E777A499EE74A7952C382D1D8471D@EUR-MSG-01.europe.corp.microsoft.com> > I was wondering how to read and write from a Socket in the most > performant way. According to the library documentation from GHC, > 'hGetBuf' and 'hPutBuf' are the way to go. But my problem with these > functions is that they expect to read exactly the number of bytes I > gave them! > > The IO computation > > hGetBuf sock ptr 1024 > > wants to read 1024 bytes, no less. This doesn't work for me, because I > need to drive an interactive line-based protocol session. I need the > read operation to consume as much data as is available and then to > return immediately, so that I can process the command right away. > > As an alternative, I looked at 'hGetLine', but this doesn't work for > me either, at least not very well, because in my protocol, lines are > terminated by "\r\n" -- not just "\n", like hGetLine expects it. hGetLine should work fine, indeed that's what I used in the Haskell Web Server. Each line will have an extra '\r' at the end, but you just need to allow for that. We really should have non-blocking versions of hGetBuf and hPutBuf, and in fact I have the code sitting in my tree. I'll take a look and see if it can be committed. Cheers, Simon From simonmar at microsoft.com Tue Sep 16 15:32:45 2003 From: simonmar at microsoft.com (Simon Marlow) Date: Tue Sep 16 09:32:57 2003 Subject: simpler I/O buffering [was: RE: An IO Question from a Newbie] Message-ID: <3429668D0E777A499EE74A7952C382D1D84782@EUR-MSG-01.europe.corp.microsoft.com> > Writing a sequence of characters across the interface I'm > proposing is a > request by the writing program for those characters to appear > at their > destination "soon". Ideally, "soon" would be "immediately"; > however, the > characters' appearance may deliberately be delayed ("buffered"), for > efficiency, as long as such delay is "unobtrusive" to a human > user of the > program. Buffering timeouts would depend on the device; for > a terminal, > perhaps 50-100 ms would be appropriate. Such an interval > would tend not > to be noticeable to a human user but would be long enough to > effectively > collect, say, an entire line of output for output "in one > piece". The use > of a reasonable timeout would avoid the confusing behavior where a > newline-less prompt doesn't appear until the prompted data is entered. So, let me see if I've understood: the situation would be exactly as now, except that the buffer automatically gets flushed if it stays non-empty for more than a certain time interval. I don't think I like it: whatever time interval you choose will be too long. If you're doing communication with another program, for example, you really don't want to have the dialog stalled for 100ms doing nothing. Sure, you could insert a flush, but only a programmer knowledgeable about buffering will know to do that, or a programmer who has debugged the program in detail to find out why it is sitting idle most of the time. So, in the end the programmer still needs to know about buffering. I'd prefer it if the program didn't work at all when the flush is missing, because it tells you immediately that something is wrong. Cheers, Simon From simons at cryp.to Tue Sep 16 16:59:10 2003 From: simons at cryp.to (Peter Simons) Date: Tue Sep 16 09:59:15 2003 Subject: Querying DNS from Haskell Message-ID: <87brtk6dht.fsf@peti.cryp.to> Hi, is anyone aware of a library for Haskell, which allows to query domain name servers for MX, A, and PTR records in a non-blocking way? My first guess was to create an Interface to ADNS [1], but I figured, maybe someone else has already solved this problem? Peter [1] http://www.gnu.org/software/adns/ From t.zielonka at students.mimuw.edu.pl Tue Sep 16 17:26:32 2003 From: t.zielonka at students.mimuw.edu.pl (Tomasz Zielonka) Date: Tue Sep 16 10:26:40 2003 Subject: Querying DNS from Haskell In-Reply-To: <87brtk6dht.fsf@peti.cryp.to> References: <87brtk6dht.fsf@peti.cryp.to> Message-ID: <20030916142631.GA31729@students.mimuw.edu.pl> On Tue, Sep 16, 2003 at 03:59:10PM +0200, Peter Simons wrote: > Hi, > > is anyone aware of a library for Haskell, which allows to query domain > name servers for MX, A, and PTR records in a non-blocking way? > > My first guess was to create an Interface to ADNS [1], but I figured, > maybe someone else has already solved this problem? I have written such a library, but it is a bit incomplete now. It is almost 100% pure Haskell (I had to use htonl/ntohl C functions in one place because I prefer to keep IP addresses in host byte order and Network.Socket.HostAddress uses network byte order). The library contains a full DNS message parser and unparser. I haven't yet implemented the full iterative query algorithm and CNAMEs are not followed automatically. I have used this library to concurrently resolve some millions of IP addresses with performance of about 50 addresses per second. If you are interested, I could clean it a bit and send you the library with an example program. > Peter Best regards, Tom -- .signature: Too many levels of symbolic links From Jon.Fairbairn at cl.cam.ac.uk Wed Sep 17 17:43:42 2003 From: Jon.Fairbairn at cl.cam.ac.uk (Jon Fairbairn) Date: Wed Sep 17 11:43:48 2003 Subject: Improvement on this function In-Reply-To: Your message of "Wed, 17 Sep 2003 10:31:11 EDT." <1063809070.9361.3.camel@hanta.promisemark.com> Message-ID: <11902.1063813422@cl.cam.ac.uk> On 2003-09-17 at 10:31EDT Gordon James Miller wrote: Something I think is more caf? than language: > Hello all. > > I'd be interested in getting some feedback on how to do this linear > interpolation function a little more cleanly. The normal way that this > is taught is to find the set of indices in the x list that bracket the > input value, then use the slope between these points to calculate the y > value for the input value. > > I had a version working at one point using !! to access particular > elements but I wasn't sure which one was the better solution. !! can be expensive. > linterp :: [Double] -> [Double] -> Double -> Double > linterp (x1:x2:xs) (y1:y2:ys) x > | x <= x2 || xs == [] = linterpPair x1 x2 y1 y2 x > | otherwise = linterp (x2:xs) (y2:ys) x > where linterpPair x1 x2 y1 y2 x = x1 + (x - x1) * (y2 - y1) / (x2 - > x1) It seems to me that you have too much going on in one function. It would be better to break it into the part that finds the point where the interpolation is supposed to happen and the function that does the interpolation. You sort-of do this, but don't go far enough for my taste. You don't say whether the xs are supposed to be in increasing order. Also it's not obvious that your version does the right thing in the case where x < head xs, and linterpPair doesn't look right to me. Something like: linterp :: [Double] -> [Double] -> Double -> Double linterp xs ys x = linterpPair (neighbours (zip xs ys)) where linterpPair ((x1,y1),(x2,y2)) = y1 + (x - x1) * (y2 - y1) / (x2 - x1) neighbours all@(p@(x1,y1):rest) | x >= x1 = head (dropWhile (( Double -> Double since that seems more natural as the data are connected that way, and it would eliminate a zip. -- J?n Fairbairn Jon.Fairbairn@cl.cam.ac.uk From simons at cryp.to Wed Sep 17 19:48:14 2003 From: simons at cryp.to (Peter Simons) Date: Wed Sep 17 12:48:18 2003 Subject: Fast I/O with sockets References: <3429668D0E777A499EE74A7952C382D1D8471D@EUR-MSG-01.europe.corp.microsoft.com> Message-ID: <877k47gy41.fsf@peti.cryp.to> Simon Marlow writes: > hGetLine should work fine, indeed that's what I used in the Haskell > Web Server. Well, another -- and more important -- problem with hGetLine is that it will read an arbitrarily long line. An attacker might use this to implement an denial-of-service attack simply by sending excessively long lines, thus overflowing the stack / heap. IMHO asynchronous hGetBuf / hPutBuf functions would really be the best solution. (Or a way to set a read timeout for hGetContents ...) Peter From gk at ninebynine.org Thu Sep 18 14:44:07 2003 From: gk at ninebynine.org (Graham Klyne) Date: Thu Sep 18 09:03:09 2003 Subject: (Off-topic) Question about categories Message-ID: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> (I'm asking here because I believe there is some overlap between category theory and functional programming cognoscenti...) It has been suggested to me that categories may be a useful framework (more useful than set theory) for conceptualizing things for which there is no well-defined equality relationship. (The discussion is about resources in WWW.) I've done a little digging about category theory, and find some relatively approachable material at: http://www.wikipedia.org/wiki/Category_theory http://www.wikipedia.org/wiki/Class_(set_theory) and nearby. But I'm hitting a mental block with this (and other places I've look don't add anything I can grok). The definition of a category depends on the definition of a morphism, and in particular the existence of an identity morphism for every object in a category. The definition of an morphism is in terms of equality of compositions of morphisms: for f : A -> B we have Id[B]. f = f = f . Id[A] My problem is this: how does it make sense to define an equality of morphisms without some well-defined concept of equality on the underlying objects to which they apply? That is, given object X and an object Y, it is possible to examine them and determine whether or not they are the same object. And if the underlying objects have such a concept of equality, what prevents them from being sets or members of sets? But categories are presented as being more general than sets. Does anyone see the cause of my non-comprehension here? #g ------------ Graham Klyne GK@NineByNine.org From franka at cs.uu.nl Thu Sep 18 16:53:30 2003 From: franka at cs.uu.nl (Frank Atanassow) Date: Thu Sep 18 09:53:34 2003 Subject: (Off-topic) Question about categories In-Reply-To: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> Message-ID: <7C712576-E9DF-11D7-9F1B-0003936416C0@cs.uu.nl> On donderdag, sep 18, 2003, at 14:44 Europe/Amsterdam, Graham Klyne wrote: > My problem is this: how does it make sense to define an equality of > morphisms without some well-defined concept of equality on the > underlying objects to which they apply? That is, given object X and > an object Y, it is possible to examine them and determine whether or > not they are the same object. Yes; there is an equality relation on the objects. The collection of objects of a (small) category forms a set. You can only form sets of things which satisfy an equality relation, because the extensionality axiom, which defines when two sets are equal, is defined in terms of equality of members. By foundation, there has to be an equality relation on the individuals in any such universe of sets. > And if the underlying objects have such a concept of equality, what > prevents them from being sets or members of sets? I don't understand your confusion. The objects can certainly be sets; the category of sets and functions (which is not small, BTW) is the archetypal example. To determine if two sets are equal you use extensionality: sets A and B are equal iff they have equal members. Nothing prevents forming a set of objects either. > But categories are presented as being more general than sets. They're more general in the sense that: a) the collection of objects can form a proper class, so a large discrete (no non-identity arrows) category is "bigger" than any set; b) the definition of a category involves more data than the definition of a set. To define a set, you only need to specify its elements. To define a category you need to give a collection of objects, an indexed collection of homsets and a composition operator. c) the collection of all sets forms one particular category. > Does anyone see the cause of my non-comprehension here? AFAICT, the source of your confusion is that you erroneously believed that sets and/or members of sets couldn't (necessarily) be compared for equality, but the fact is they always can, by definition. If the category is large, there are additional issues, but there is still an equality relation on the objects. Does that clear things up? Regards, Frank From alastair at reid-consulting-uk.ltd.uk Thu Sep 18 18:02:06 2003 From: alastair at reid-consulting-uk.ltd.uk (Alastair Reid) Date: Thu Sep 18 12:02:10 2003 Subject: (Off-topic) Question about categories In-Reply-To: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> References: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> Message-ID: <200309181702.06770.alastair@reid-consulting-uk.ltd.uk> > But I'm hitting a mental block with this (and other places I've > look don't add anything I can grok). Two things I found useful when learning category theory are: 1) Try reading the definitions as though categories were just graphs: Objects are nodes, morphisms are labelled edges and equality of morphism is modulo some equivalence on lists of labels. (Well, something like that - it's been a while since I played this game and it's most useful for you to work out all the ins and outs of it yourself.) 2) Look for lots and lots of examples of categories. Besides the obvious category of sets, there's domain theory (very useful to help you get a handle on pushouts, initial, final, and the like), and lots and lots of others. (Ask computer scientists for examples though - if you have to learn topology before you can understand the categories that mathematicians like to use, you won't gain much insight.) Actually, these are just two different ways of saying to look beyond the obvious Set category. The value of category theory is in its generalizations and, especially, in its ability to draw parallels between many different situations so looking at any one example for too long will tend to hold you back. -- Alastair Reid www.haskell-consulting.com From diatchki at cse.ogi.edu Thu Sep 18 11:48:13 2003 From: diatchki at cse.ogi.edu (Iavor Diatchki) Date: Thu Sep 18 13:48:19 2003 Subject: (Off-topic) Question about categories In-Reply-To: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> References: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> Message-ID: <3F69EFDD.1080504@cse.ogi.edu> hello, i think what is importnat is that one can tell when two arrows are the same (how to specify that of course is tricky in general). equality on objects kind of follows from there. in fact one can describe a category without referring to objects at all: 1. pick a collection of arrows (call it A) 2. pick a sub-collection if A that you will call "identity" arrows (call that I) 3. give the following mappings (these are not arrows!) 3.1 source: A -> I 3.2 target: A -> I 3.3 compose: (A,A) -> A (this one is partial, and has to be defined, when the source and target match as usual). the trick is that essentially objects are represented by the identity arrows on them (as there is a 1-1 correspondance between the two). my favourite way to think of a category is as follows: 1. i think of objects as types 2. i think of an arrow: A -> B as value of type B, that has a free variable of type A 3. i think of composition as substitution hope this helps iavor Graham Klyne wrote: > (I'm asking here because I believe there is some overlap between > category theory and functional programming cognoscenti...) > > It has been suggested to me that categories may be a useful framework > (more useful than set theory) for conceptualizing things for which there > is no well-defined equality relationship. (The discussion is about > resources in WWW.) I've done a little digging about category theory, > and find some relatively approachable material at: > http://www.wikipedia.org/wiki/Category_theory > http://www.wikipedia.org/wiki/Class_(set_theory) > and nearby. > > But I'm hitting a mental block with this (and other places I've look > don't add anything I can grok). The definition of a category depends on > the definition of a morphism, and in particular the existence of an > identity morphism for every object in a category. The definition of an > morphism is in terms of equality of compositions of morphisms: > for f : A -> B we have Id[B]. f = f = f . Id[A] > > My problem is this: how does it make sense to define an equality of > morphisms without some well-defined concept of equality on the > underlying objects to which they apply? That is, given object X and an > object Y, it is possible to examine them and determine whether or not > they are the same object. And if the underlying objects have such a > concept of equality, what prevents them from being sets or members of > sets? But categories are presented as being more general than sets. > > Does anyone see the cause of my non-comprehension here? > > #g > > > ------------ > Graham Klyne > GK@NineByNine.org > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > -- ================================================== | Iavor S. Diatchki, Ph.D. student | | Department of Computer Science and Engineering | | School of OGI at OHSU | | http://www.cse.ogi.edu/~diatchki | ================================================== From veger at cistron.nl Thu Sep 18 20:51:16 2003 From: veger at cistron.nl (Peter J. Veger) Date: Thu Sep 18 13:51:22 2003 Subject: (Off-topic) Question about categories In-Reply-To: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> Message-ID: <000001c37e0d$757b10e0$0a02a8c0@veger3> Graham Klyne wrote, on 18 september 2003 14:44 [...] > But I'm hitting a mental block with this (and other places I've look don't > add anything I can grok). The definition of a category depends on the > definition of a morphism, and in particular the existence of an identity > morphism for every object in a category. The definition of an morphism is > in terms of equality of compositions of morphisms: > for f : A -> B we have Id[B]. f = f = f . Id[A] > My problem is this: how does it make sense to define an equality of > morphisms without some well-defined concept of equality on the underlying > objects to which they apply? That is, given object X and an object Y, it > is possible to examine them and determine whether or not they are the same > object. And if the underlying objects have such a concept of equality, > what prevents them from being sets or members of sets? But categories are > presented as being more general than sets. I am not a specialist, but my understanding is as follows: Category Theory is an axiomatic theory: a language with certain primitive notions and relations between these notions. Equality between objects is not a notion of category theory and you will not find any category-theoretic theorem proving that two objects are equal. (You may say that an object has an identity, as given by its identity morphism. A category actually can be defined using morphisms only.) Equality between morphisms however is a central notion of category theory. If you have a concrete category, you have to denote some things as objects, others as morphisms; before exploiting category theory, you have to prove that the objects and morphisms of your concrete category fulfill the category axioms, and for that you may, of course, use the equality relations in your concrete category. Peter J. Veger, Best Netherlands From lektu at terra.es Thu Sep 18 21:16:33 2003 From: lektu at terra.es (Juanma Barranquero) Date: Thu Sep 18 14:21:31 2003 Subject: How to detect finite/infinite lists? Message-ID: <20030918200621.4D5B.LEKTU@terra.es> Extremely-newbie questions: Is there any way to know if a list is finite or infinite, other than doing: length l and waiting forever? :) I ask because I was learning Haskell by writing some pretty naive implementation of surreal numbers, where I used lists for left and right surreal sets, and I wanted to do some operations on the sets (like finding the maximum/minimum), but obviously just on finite sets. I vaguely suspect the answer is going to be: "No, because lists are lazy (at least when they are :) and there's no general way to know beforehand how many elements they're going to have". But still, if I write x = [1..5] the compiler knows pretty well x is not going to grow any new member... (Unrelated) Is there any standard function to do: interleave [] _ = [] interleave _ [] = [] interleave (x:xs) (y:ys) = x : y : interleave xs ys It's pretty easy to implement as shown or via zipWith, but given that Haskell 98 already includes some "basic" functions (like cycle, iterate, etc.) I wonder if I've missed this one somehow. Thanks, /L/e/k/t/u From brandon at its.caltech.edu Thu Sep 18 13:28:41 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Thu Sep 18 15:30:06 2003 Subject: How to detect finite/infinite lists? In-Reply-To: <20030918200621.4D5B.LEKTU@terra.es> References: <20030918200621.4D5B.LEKTU@terra.es> Message-ID: On Thu, 18 Sep 2003, Juanma Barranquero wrote: > Extremely-newbie questions: > > Is there any way to know if a list is finite or infinite, other than > doing: > > length l > > and waiting forever? :) > > I ask because I was learning Haskell by writing some pretty naive > implementation of surreal numbers, where I used lists for left and right > surreal sets, and I wanted to do some operations on the sets (like > finding the maximum/minimum), but obviously just on finite sets. > > I vaguely suspect the answer is going to be: "No, because lists are lazy > (at least when they are :) and there's no general way to know beforehand > how many elements they're going to have". But still, if I write > > x = [1..5] > > the compiler knows pretty well x is not going to grow any new member... Well, it's easy to tell that a list is finite by running length and having it terminate. This is obviously equivalent to the halting problem so you can't really expect anything better in general. Why do you need to test whether lists are infinite? If your lists are being generated from finite descriptions maybe you could use a data structure that records the descriptions. For example, rather than defining l = [1..]++[2,3..7] you could define data EnumList = EnumFromBy Integer Integer | EnumFromToBy Int Int Int | Append EnumList EnumList l = EnumFromBy 1 1 `Append` EnumFromToBy 2 1 7 then to test whether a list is infinite you can write infinite (EnumFromBy _ _) = True infinite (EnumFromToBy a delta b) = compare a delta == compare a b infinite (Append l1 l2) = infinite l1 && infinite l2 Of course this only works if it is computable whether a description gives a finite list. > (Unrelated) Is there any standard function to do: > > interleave [] _ = [] > interleave _ [] = [] > interleave (x:xs) (y:ys) = x : y : interleave xs ys > > It's pretty easy to implement as shown or via zipWith, but given that > Haskell 98 already includes some "basic" functions (like cycle, iterate, > etc.) I wonder if I've missed this one somehow. > > Thanks, Not as far as I know. The standard List module doesn't define anything like that and Data.List doesn't define anything like that. What do you want to use it for? If you are looking for strict alternation you should use this defintion. If you want a nondeterministic merge as elements are computed, look at mergeIO in Control.Concurrent in the GHC libraries. Brandon From ddarius at hotpop.com Thu Sep 18 16:53:12 2003 From: ddarius at hotpop.com (Derek Elkins) Date: Thu Sep 18 15:55:20 2003 Subject: How to detect finite/infinite lists? In-Reply-To: <20030918200621.4D5B.LEKTU@terra.es> References: <20030918200621.4D5B.LEKTU@terra.es> Message-ID: <20030918155312.000056c1.ddarius@hotpop.com> On Thu, 18 Sep 2003 20:16:33 +0200 Juanma Barranquero wrote: > Extremely-newbie questions: > > Is there any way to know if a list is finite or infinite, other than > doing: > > length l > > and waiting forever? :) In Haskell 98, no. With a slightly impure extension (observable sharing) sometimes but in general, no. The only infinity you need to recognize is omega and that's just an infinite (cyclic) list of whatevers. So you -could- use observable sharing for this, but there isn't much point to it, just use a data structure that says, "an infinity of x". The simplest thing I would think of is to follow the arithmetic operation exactly. data SN = Zero | Up | Down | SN :+: SN | SN :*: SN | SN :^: SN | Omega SN omega = Omega Up iota = Up :+: Omega Down twoThirds = Omega (Up :+: Down) twoOmega = Omega Up :+: Omega Up omegaSquared = Omega Up :*: Omega Up omegaToTheOmega = Omega Up :^: Omega Up From lektu at terra.es Thu Sep 18 22:58:09 2003 From: lektu at terra.es (Juanma Barranquero) Date: Thu Sep 18 15:58:15 2003 Subject: How to detect finite/infinite lists? In-Reply-To: <055401c37e11$97aa4c90$6401a8c0@mauve> References: <20030918200621.4D5B.LEKTU@terra.es> <055401c37e11$97aa4c90$6401a8c0@mauve> Message-ID: <20030918215305.4D61.LEKTU@terra.es> On Thu, 18 Sep 2003 14:20:46 -0400, "David Barton" wrote: > that's halting problem equivalent. Yes, I know it's a halting problem, in the general case. But the compiler could try to say if it knows it *is* a finite list. So: x = [1..10] y = [1..] z = someFunctionOrOther knownToBeFinite x -> should return Just True knownToBeFinite y -> should return Just False knownToBeFinite z -> should return Nothing ("I don't know"). or something along these lines. > Nope. Ok. Thanks, /L/e/k/t/u From lektu at terra.es Thu Sep 18 23:41:39 2003 From: lektu at terra.es (Juanma Barranquero) Date: Thu Sep 18 16:41:41 2003 Subject: How to detect finite/infinite lists? In-Reply-To: References: <20030918200621.4D5B.LEKTU@terra.es> Message-ID: <20030918223024.4D67.LEKTU@terra.es> On Thu, 18 Sep 2003 12:28:41 -0700 (PDT), Brandon Michael Moore wrote: > This is obviously equivalent to the halting problem so you > can't really expect anything better in general. No, obviously not in general. But a human wouldn't sweat to say that [1..5] is finite and [1..] infinite. As I have no idea how the compiler is implemented, I don't know if it knows whether what it has on its hands is a fixed, short-length list (like [1..5], which I imagine the compiler would perhaps pre-generate instead of lazy-evaluating it), or one which in fact comes from some kind of generator function(s). > Why do you need to test whether lists are infinite? I had something like: data Surreal = Surreal { leftSet::[Surreal], rightSet::[Surreal] } where leftSet and rightSet could potentially contain infinite lists of Surreals. Knowing that both sets are finite would allow to simplify some numbers (in surreal numbers, it is the case that {-1|1} = {|}, for example, and being able to simplify {-1|1} to {|} can help, specially when trying to show/print the numbers). > If your lists are being generated from finite > descriptions maybe you could use a data structure that records the > descriptions. Yes, I think you're right. > What do you want to use it for? If you are looking for strict alternation you > should use this defintion. Yes, strict alternation (output, in fact). I have lists of columns and lists of separators and I interleave them while generating a line of output. Thanks, /L/e/k/t/u From lektu at terra.es Thu Sep 18 23:43:40 2003 From: lektu at terra.es (Juanma Barranquero) Date: Thu Sep 18 16:43:51 2003 Subject: How to detect finite/infinite lists? In-Reply-To: <20030918155312.000056c1.ddarius@hotpop.com> References: <20030918200621.4D5B.LEKTU@terra.es> <20030918155312.000056c1.ddarius@hotpop.com> Message-ID: <20030918224257.4D6A.LEKTU@terra.es> On Thu, 18 Sep 2003 15:53:12 -0400, Derek Elkins wrote: > In Haskell 98, no. With a slightly impure extension (observable > sharing) sometimes but in general, no. Interesting. > just use a data structure that says, "an > infinity of x". The simplest thing I would think of is to follow the > arithmetic operation exactly. > > data SN > = Zero > | Up > | Down > | SN :+: SN > | SN :*: SN > | SN :^: SN > | Omega SN :) Thanks, /L/e/k/t/u From karczma at info.unicaen.fr Fri Sep 19 11:44:24 2003 From: karczma at info.unicaen.fr (Jerzy Karczmarczuk) Date: Fri Sep 19 04:44:28 2003 Subject: Is this a terminologic issue? References: <20030918200621.4D5B.LEKTU@terra.es> <20030918223024.4D67.LEKTU@terra.es> Message-ID: <3F6AC1E8.9060005@info.unicaen.fr> [[Sent to Haskell-caf?, and to comp.lang.fun]] If you look at some Web sites (Mathworld, the site of John Baez - a known spec. in algebraic methods in physics), or into some books on differential geometry, you might easily find something which is called pullback or pull-back. Actually, it is the construction of a dual, whose meaning can be distilled and implemented in Haskell as follows. The stuff is very old, and very well known. Suppose you have two domains X and Y. A function F : X -> Y. The form (F x) gives some y. You have also a functor which constructs the dual spaces, X* and Y* - spaces of functionals over X or Y. A function g belongs to Y* if g : Y -> Z (some Z, let's keep one space like this). Now, I can easily construct a dual to F, the function F* : Y* -> X* by (F* g) x = g (F x) and this mapping is called pullback... While there is nothing wrong with that, and in Haskell one may easily write the 'star' generator (star f) g x = g (f x) or star = flip (.) ... I have absolutely no clue why this is called a pullback. Moreover, in the incriminated diff. geom. books, its inverse is *not* called pushout, but push-forward. Anyway, I cannot draw any pullback diagram from that. The closest thing I found is the construction in Asperti & Longo, where a F in C[A,B] induces F* : C!B -> C!A where the exclam. sign is \downarrow, the "category over ...". The diagram is there, a 9-edge prism, but - in my eyes - is quite different from what one can get from this "contravariant composition" above. But my eyes are not much better than my ears, so... I sent this question to a few gurus, and the answers are not conclusive, although it seems that this *is* a terminologic confusion. Vincent Danos wrote: > it really doesn't look like a categorical pullback > and it might well be a "pull-back" only in the sense > that if if F:A->B is a linear map say and f is a linear form on B, then F*(f) > is a linear form on A > defined as F*(f)(a)=f(b=F(a)) so one can "pull back" (linearly of course!) > linear forms on B to linear forms on A > "back" refers to the direction of F, i'd say. ================================== Does anybody have a different (or any!) idea about that? Thank you in advance for helping me to solve my homework. Jerzy Karczmarczuk Caen, France From daniel_yokomiso at yahoo.com.br Fri Sep 19 09:01:12 2003 From: daniel_yokomiso at yahoo.com.br (Daniel Yokomiso) Date: Fri Sep 19 06:55:59 2003 Subject: (Off-topic) Question about categories References: <5.1.0.14.2.20030918132357.02fdbf30@127.0.0.1> Message-ID: <030001c37e9d$86dcd310$0100000a@eonl9byru76yj2> ----- Original Message ----- From: "Graham Klyne" To: "Haskell Cafe" Sent: Thursday, September 18, 2003 9:44 AM Subject: (Off-topic) Question about categories > (I'm asking here because I believe there is some overlap between category > theory and functional programming cognoscenti...) > > It has been suggested to me that categories may be a useful framework (more > useful than set theory) for conceptualizing things for which there is no > well-defined equality relationship. (The discussion is about resources in > WWW.) I've done a little digging about category theory, and find some > relatively approachable material at: > http://www.wikipedia.org/wiki/Category_theory > http://www.wikipedia.org/wiki/Class_(set_theory) > and nearby. > > But I'm hitting a mental block with this (and other places I've look don't > add anything I can grok). The definition of a category depends on the > definition of a morphism, and in particular the existence of an identity > morphism for every object in a category. The definition of an morphism is > in terms of equality of compositions of morphisms: > for f : A -> B we have Id[B]. f = f = f . Id[A] > > My problem is this: how does it make sense to define an equality of > morphisms without some well-defined concept of equality on the underlying > objects to which they apply? That is, given object X and an object Y, it > is possible to examine them and determine whether or not they are the same > object. And if the underlying objects have such a concept of equality, > what prevents them from being sets or members of sets? But categories are > presented as being more general than sets. > > Does anyone see the cause of my non-comprehension here? > > #g > > > ------------ > Graham Klyne > GK@NineByNine.org Hi, There's some additional links about category theory in this LtU discussion: http://lambda.weblogs.com/discuss/msgReader$979. Also there's this presentation "Category Theory for Beginners" http://www.cs.toronto.edu/~sme/presentations/cat101.pdf. HTH. Best regards, Daniel Yokomiso. "Honesty is the best policy, but insanity is a better defense." --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.518 / Virus Database: 316 - Release Date: 12/9/2003 From GK at ninebynine.org Mon Sep 22 14:28:48 2003 From: GK at ninebynine.org (Graham Klyne) Date: Mon Sep 22 08:35:00 2003 Subject: Thanks: (was Question about categories) Message-ID: <5.1.0.14.2.20030922130621.00bbab10@127.0.0.1> Thanks to everyone who responded to my question about category theory. The responses turned up some interesting and useful material on the web, which I'm still digesting. Notably [1] [2]. There's something in the introduction of [2] that I think will help me get a handle on this: [[ ... objects are not collections of "elements," and morphisms do not need to be functions between sets (thus morphisms cannot be applied to "elements" but only composed with other morphisms). Any immediate access to the internal structure of objects is prevented: all properties of objects must be specified by properties of morphisms ... ]] My earlier supposition was that an "object" was roughly in correspondence with a member of a set, but I'm coming to perceive that "object" corresponds more with the notion of a set or collection of some (unspecified) things. #g -- [1] http://www.cs.toronto.edu/~sme/presentations/cat101.pdf [2] http://www.di.ens.fr/users/longo/download.html which has a link to content of the (apparently) out-of-print book: [[ CATEGORIES TYPES AND STRUCTURES An Introduction to Category Theory for the working computer scientist Andrea Asperti Giuseppe Longo FOUNDATIONS OF COMPUTING SERIES, M.I.T. PRESS, 1991 ]] _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ------------ Graham Klyne GK@NineByNine.org From jcast at cate0-46.reshall.ou.edu Sun Sep 21 22:02:57 2003 From: jcast at cate0-46.reshall.ou.edu (Jon Cast) Date: Mon Sep 22 08:57:27 2003 Subject: An IO Question from a Newbie In-Reply-To: Message from Derek Elkins "of Fri, 12 Sep 2003 19:17:05 EDT." <20030912191705.0000677a.ddarius@hotpop.com> References: <200309121627.59788.angagon@earthlink.net> <20030912191705.0000677a.ddarius@hotpop.com> Message-ID: > On Fri, 12 Sep 2003 16:27:59 -0600 > Matt O'Connor wrote: > > > Hello all. I'm new to functional programming and Haskell, but have > > been programming in C and Java for a while. I've been going through > > the tutorials and whatnot on haskell.org. I've read from the Gentle > > Introduction to Haskell about IO and some of the other stuff and I > > have a question about it. > > > > main = do putStr "Type Something: " > > str <- getLine > > putStrLn ("You typed: " ++ str) > > > > When compile and run this code the "Type Something: " isn't displayed > > until the putStrLn. So there is no prompt. The output looks like > > this. > > > > s > > Type Something: You typed: s > > > > But if I change the putStr "Type Something: " to a putStrLn or put a > > \n at the end of the string it displays the text immediately (ie, when > > I want it to). Is there a good reason for this? Am I doing something > > wrong? Or do I need to call some kind of standard output flush? > > Yes, the problem is that the first line is not being flushed. Importing > hFlush and stdout from IO (or System.IO) and adding hFlush stdout after > you putStr will fix it. Alternatively, you could change the buffering > on stdout to NoBuffering. I don't know if there is a standard > putStrWithFlush function or just a better answer than this. A > putStrWithFlush would be trivial to write or more generally to a > HOF that flushed after performing the action passed to it. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Just thought I'd point out (in case it isn't clear) that this function is flushAfter h a = liftM2 const a (hFlush h). Not much of an improvement... Jon Cast From Alistair_Bayley at ldn.invesco.com Mon Sep 22 13:10:36 2003 From: Alistair_Bayley at ldn.invesco.com (Bayley, Alistair) Date: Mon Sep 22 08:57:28 2003 Subject: garbage collection and Ptr/ForeignPtr/StablePtr Message-ID: <7DFF3BC6CA957441AEADA7F340BFAA3401B33735@GBLONEX11.lon.invesco.com> Should I expect that Ptr memory allocated with malloc is garbage collected when no longer used by the Haskell runtime? The FFI spec doesn't say so explicitly (that I could see); AFAICT this is how the pointer types behave: Ptr : freed when no longer used/referenced ForeignPtr : like Ptr, but will also run a finaliser when freed StablePtr : not freed or collected (I haven't tested this; I've only read the FFI spec.) ***************************************************************** The information in this email and in any attachments is confidential and intended solely for the attention and use of the named addressee(s). This information may be subject to legal professional or other privilege or may otherwise be protected by work product immunity or other legal rules. It must not be disclosed to any person without our authority. If you are not the intended recipient, or a person responsible for delivering it to the intended recipient, you are not authorised to and must not disclose, copy, distribute, or retain this message or any part of it. ***************************************************************** From simonmar at microsoft.com Mon Sep 22 15:11:04 2003 From: simonmar at microsoft.com (Simon Marlow) Date: Mon Sep 22 09:09:44 2003 Subject: garbage collection and Ptr/ForeignPtr/StablePtr Message-ID: <3429668D0E777A499EE74A7952C382D1DFE276@EUR-MSG-01.europe.corp.microsoft.com> > Should I expect that Ptr memory allocated with malloc is > garbage collected > when no longer used by the Haskell runtime? The FFI spec > doesn't say so > explicitly (that I could see); AFAICT this is how the pointer > types behave: > Ptr : freed when no longer used/referenced > ForeignPtr : like Ptr, but will also run a finaliser when freed > StablePtr : not freed or collected No, all memory that you allocate with the FFI must be freed explicitly. You can arrange for memory to be deallocated when no longer referenced by using a ForeignPtr, and using the finalizer to free the memory. Bear in mind that if your application has large, or strict, memory requirements, then relying on the ForeignPtr method may not be the best solution, because finalizers can be unpredictable. Cheers, Simon From daniel_yokomiso at yahoo.com.br Mon Sep 22 14:38:00 2003 From: daniel_yokomiso at yahoo.com.br (=?iso-8859-1?q?Daniel=20Yokomiso?=) Date: Mon Sep 22 12:38:03 2003 Subject: Thanks: (was Question about categories) In-Reply-To: <5.1.0.14.2.20030922130621.00bbab10@127.0.0.1> Message-ID: <20030922163800.30111.qmail@web40603.mail.yahoo.com> --- Graham Klyne escreveu: [snip] > [2] http://www.di.ens.fr/users/longo/download.html > which has a link to content of the (apparently) > out-of-print book: > [[ > CATEGORIES TYPES AND STRUCTURES > An Introduction to Category Theory for the working > computer scientist > Andrea Asperti > Giuseppe Longo > FOUNDATIONS OF COMPUTING SERIES, M.I.T. PRESS, 1991 > ]] Seems to be this book: http://www.amazon.com/exec/obidos/tg/detail/-/0262011255/qid=1064248079/sr=8-1/ref=sr_8_1/002-2181686-7144062?v=glance&s=books&n=507846 There's also a list of some books about it: http://www.booksunderreview.com/Science/Math/Algebra/Category_Theory/ BTW I tried to connect to ftp.ens.fr/pub/dmi/users/longo/CategTypesStructures (as informed in an Amazon review) but failed. Someone knows of another place where we can legally obtain the postscript files (if they exist at all)? Best regards, Daniel Yokomiso. _______________________________________________________________________ Desafio AntiZona: participe do jogo de perguntas e respostas que vai dar um Renault Clio, computadores, c?meras digitais, videogames e muito mais! www.cade.com.br/antizona From diatchki at cse.ogi.edu Mon Sep 22 12:27:53 2003 From: diatchki at cse.ogi.edu (Iavor Diatchki) Date: Mon Sep 22 14:27:57 2003 Subject: Thanks: (was Question about categories) In-Reply-To: <20030922163800.30111.qmail@web40603.mail.yahoo.com> References: <20030922163800.30111.qmail@web40603.mail.yahoo.com> Message-ID: <3F6F3F29.20708@cse.ogi.edu> hi, for the book mentioned bellow (and many other useful texts) check out frank atanassow page: http://www.cs.uu.nl/people/franka/ref hope this helps iavor Daniel Yokomiso wrote: > --- Graham Klyne escreveu: > > [snip] > > >>[2] http://www.di.ens.fr/users/longo/download.html >>which has a link to content of the (apparently) >>out-of-print book: >>[[ >>CATEGORIES TYPES AND STRUCTURES >>An Introduction to Category Theory for the working >>computer scientist >>Andrea Asperti >>Giuseppe Longo >>FOUNDATIONS OF COMPUTING SERIES, M.I.T. PRESS, 1991 >>]] > > > Seems to be this book: > > http://www.amazon.com/exec/obidos/tg/detail/-/0262011255/qid=1064248079/sr=8-1/ref=sr_8_1/002-2181686-7144062?v=glance&s=books&n=507846 > > There's also a list of some books about it: > > http://www.booksunderreview.com/Science/Math/Algebra/Category_Theory/ > > BTW I tried to connect to > ftp.ens.fr/pub/dmi/users/longo/CategTypesStructures > (as informed in an Amazon review) but failed. Someone > knows of another place where we can legally obtain the > postscript files (if they exist at all)? > > Best regards, > Daniel Yokomiso. > > > _______________________________________________________________________ > Desafio AntiZona: participe do jogo de perguntas e respostas que vai > dar um Renault Clio, computadores, c?meras digitais, videogames e muito > mais! www.cade.com.br/antizona > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe@haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > -- ================================================== | Iavor S. Diatchki, Ph.D. student | | Department of Computer Science and Engineering | | School of OGI at OHSU | | http://www.cse.ogi.edu/~diatchki | ================================================== From bhalchin at hotmail.com Mon Sep 22 21:38:12 2003 From: bhalchin at hotmail.com (Bill Halchin) Date: Mon Sep 22 16:38:15 2003 Subject: Thanks: (was Question about categories) Message-ID: An HTML attachment was scrubbed... URL: http://porky.devel.redhat.com/pipermail/haskell-cafe/attachments/20030922/6b3dcbf0/attachment.htm From florian at hars.de Tue Sep 23 09:37:24 2003 From: florian at hars.de (Florian Hars) Date: Tue Sep 23 02:37:37 2003 Subject: Thanks: (was Question about categories) In-Reply-To: References: Message-ID: <3F6FEA24.2050908@hars.de> Bill Halchin wrote: > Here is the book plus many other llongo papers: > http://www.di.ens.fr/users/llongo/download.html s/llongo/longo/ From alastair at reid-consulting-uk.ltd.uk Tue Sep 23 13:11:01 2003 From: alastair at reid-consulting-uk.ltd.uk (Alastair Reid) Date: Tue Sep 23 07:11:07 2003 Subject: garbage collection and Ptr/ForeignPtr/StablePtr In-Reply-To: <7DFF3BC6CA957441AEADA7F340BFAA3401B33735@GBLONEX11.lon.invesco.com> References: <7DFF3BC6CA957441AEADA7F340BFAA3401B33735@GBLONEX11.lon.invesco.com> Message-ID: <200309231211.01984.alastair@reid-consulting-uk.ltd.uk> On Monday 22 September 2003 12:10 pm, Bayley, Alistair wrote: > Should I expect that Ptr memory allocated with malloc is garbage collected > when no longer used by the Haskell runtime? The FFI spec doesn't say so > explicitly (that I could see); C programs use pointers in many ways: pointers to static objects, pointers to stack-allocated objects, etc. as well as pointers to malloc-allocated objects. These are all represented with a 'Ptr' in Haskell but only one of them should be freed. Also, there are many ways to free a malloced object in C programs. You might have to decrement a reference count, you might have to first free any objects it contains pointers to, you might call a library-specific free routine called 'xfree', you might call a type-specific freeing routine or you might just call free. For all these reasons, Haskell's garbage collector cannot just call 'free' for you. > AFAICT this is how the pointer types behave: > Ptr : freed when no longer used/referenced No. If it needs freed, you have to explicitly deallocate this by calling an appropriate routine. This effectively means that you use Ptr for things accessed inside the IO monad. > ForeignPtr : like Ptr, but will also run a finaliser when freed Yes. The finalizer is (a pointer to) any C function you like. ForeignPtr is especially useful for datatypes accessed outside the IO monad. e.g., if you wanted to use a complex number library, you might use ForeignPtr for the freeing function. ForeignPtr is also useful even if you are in the IO monad when you don't have a simple access pattern like 'allocate; use; free;'. Note that the ffi libraries provide functions called 'malloc' and 'free' and a finalizer called 'finalizerFree'. It is tempting to think that these are the C library 'malloc'/'free' and (a pointer to) 'free'. This is not necessarily so and you must take care that: - you only use MarshallAlloc.free and MarshallAlloc.finalizerFree with memory allocated using MarshallAlloc.malloc - you only use libc's 'free' with memory allocated using libc's 'malloc'. > StablePtr : not freed or collected Yes. StablePtrs are in a different category because they are pointers from C into Haskell (in contrast, Ptrs and ForeignPtrs are pointers from Haskell into C). Since C doesn't have garbage collection, it is up to your C code to explicitly call 'freeStablePtr'. > (I haven't tested this; I've only read the FFI spec.) I had a quick reread of sections 5.4 and 5.5 and I think your confusion is understandable. It doesn't say anything about deallocation so you might plausibly guess that the story is the same as for Int8 (say) which also says nothing about deallocation. ForeignPtrs are contrasted with Ptrs but only in whether or not they have finalizers. I think the ffi spec has now been finalized so it's too late to fix this but perhaps the next revision could include the following wording (inserted after the first sentence of section 5.4). Objects of type 'Ptr a' and 'FunPtr a' are garbage collected in the normal way but the objects they _point to_ are not garbage collected. If the objects pointed to require deallocation, this must be done by explicitly calling an appropriate function such as the C library function 'free'. I have put some of the above in my FFI guide at: http://www.reid-consulting-uk.ltd.uk/docs/ffi.html -- Alastair Reid www.haskell-consulting.com From simons at cryp.to Tue Sep 23 16:06:19 2003 From: simons at cryp.to (Peter Simons) Date: Tue Sep 23 09:06:24 2003 Subject: Generating "setMember" functions for record structures Message-ID: <87wubzr6wk.fsf@peti.cryp.to> Hi, in one of my programs, I have to modify individual fields of a pretty large record structure quite frequently. Assume the structure is: > data State = State { foo :: [Int] > , bar :: [String] > } Then I'd like to have functions setFoo and setBar, which copy a State but replace the field in question with a given value. This is implemented as: > setFoo :: State -> [Int] -> State > setFoo st x = State { foo = x > , bar = bar st > } > setBar :: State -> [String] -> State > setBar st x = State { foo = foo st > , bar = x > } Now, defining these helper functions for large records is not what I call an interesting job. Thus I wonder, is there any way to automate this task? I thought about using Template Haskell, but I haven't used it at all yet and am uncertain about how difficult accomplishing this would be. I also thought about using "Data.Generics", which I know better, but the generic approach feels like over-kill for this task. In order to use it, I'd have to re-design my data structure as well. So this isn't too attractive ... Any ideas? Comments? Source Code? :-) Peter From maeder at tzi.de Tue Sep 23 16:38:16 2003 From: maeder at tzi.de (Christian Maeder) Date: Tue Sep 23 09:38:26 2003 Subject: Generating "setMember" functions for record structures In-Reply-To: <87wubzr6wk.fsf@peti.cryp.to> References: <87wubzr6wk.fsf@peti.cryp.to> Message-ID: <3F704CC8.4020105@tzi.de> >>setFoo :: State -> [Int] -> State >>setFoo st x = State { foo = x >> , bar = bar st >> } > > >>setBar :: State -> [String] -> State >>setBar st x = State { foo = foo st >> , bar = x >> } It should be sufficient to just write: setFoo st x = st { foo = x } setBar st x = st { bar = x } This creates a new State (out of "st") with an updated field entry without mentioning the other fields. Christian From Malcolm.Wallace at cs.york.ac.uk Tue Sep 23 15:35:51 2003 From: Malcolm.Wallace at cs.york.ac.uk (Malcolm Wallace) Date: Tue Sep 23 09:40:24 2003 Subject: Generating "setMember" functions for record structures In-Reply-To: <87wubzr6wk.fsf@peti.cryp.to> References: <87wubzr6wk.fsf@peti.cryp.to> Message-ID: <20030923143551.78611cb3.Malcolm.Wallace@cs.york.ac.uk> Peter Simons writes: > > setFoo :: State -> [Int] -> State > > setFoo st x = State { foo = x > > , bar = bar st > > } There is an easier way to do this, using the record update syntax rather than record construction syntax. e.g. setFoo st x = st { foo = x } All the unmentioned fields retain their previous values from the 'st' structure. Regards, Malcolm From Alistair_Bayley at ldn.invesco.com Tue Sep 23 15:29:20 2003 From: Alistair_Bayley at ldn.invesco.com (Bayley, Alistair) Date: Tue Sep 23 10:03:20 2003 Subject: Database interface - would like advice on oracle library binding Message-ID: <7DFF3BC6CA957441AEADA7F340BFAA3401B33742@GBLONEX11.lon.invesco.com> (2nd attempt; mailman thinks I'm not a list member, but it still keeps sending me mail.) Still making slow progress on an Oracle database binding... now I'm trying to fit the API I have into some sort of abstract interface (like the one(s) discussed previously: http://haskell.org/pipermail/haskell-cafe/2003-August/004957.html ). 1. Is the left-fold the best/only interface to expose? I think yes, but that doesn't allow fine-grained control over cursors i.e. being able to open many cursors at once and interleave fetches from them. Or does it? 2. I'm finding it hard to write a doQuery function that takes an extraction function that isn't a pig to write. Some advice would be useful here... (and a long-ish explanation follows): The Oracle Call Interface (OCI) requires that I allocate buffers for the result of a single row fetch, before the first row is fetched. So a query involves: - prepare statement etc - allocate buffers (one for each column - call OCI C function "DefineByPos") - fetch row - extract/marshal data from buffer into Haskell types (which are then processed by fold function) - fetch, marshal (repeat until no more rows) - free buffers i.e. the same buffers are re-used for each row. The problem for me is how to specify the left-fold function in terms of the low-level API. If I want to supply extraction functions (to extract Haskell values from result buffer), how do I manage the buffer allocation in the doQuery function? The buffer allocate/free code also needs to know the column positions and types, in the same manner as the extract functions. I want to be able to write code like this: results <- doQuery dbconn sqltext [] \row results -> do name <- stringv row 1 address <- stringv row 2 return (name,address):results .. where the stringv function extracts/marshals a Haskell String from the result buffer. The intermediate approach I currently have means I have to pass an IO action into the doQuery function that, when evaluated, allocates the buffer and returns two more actions: - an action that extracts the row as a tuple - another action that frees the buffer The doQuery function evaluates the initial action (to allocate the buffer), uses the extract action to build the result (at present a list), and when there are no more rows, uses the free action to free the buffer. This approach is quite awkward (especially w.r.t. writing extract functions), and it's hard for me to see how to build a better interface. Hard, because of the memory management requirements. -------------------- Here's a chunk of the code. A lot of it is OCI plumbing, but I hope you can see how awkward it is to create an extract function (see ex3 at the bottom). Given pointers to the buffer, extract a string of variable length (you have to terminate it yourself). > fetchStringVal :: OCIColInfo -> IO String > fetchStringVal (_, bufptr, nullindptr, retsizeptr) = do > retsize <- liftM cShort2Int (peek retsizeptr) > nullind <- liftM cShort2Int (peek nullindptr) -- unused > pokeByteOff (castPtr bufptr) retsize nullByte > val <- peekCString (castPtr bufptr) > return val Free a single column's buffer. > freeColBuffer :: OCIColInfo -> IO () > freeColBuffer (_, bufptr, nullindptr, retsizeptr) = do > free bufptr > free retsizeptr > free nullindptr Create a buffer for a string column, and return the extract and free IO actions. > getExtractFnString :: Int -> ErrorHandle -> StmtHandle -> IO (IO String, IO ()) > getExtractFnString posn err stmt = do > c <- defineCol err stmt posn 2000 oci_SQLT_CHR > return ((fetchStringVal c), (freeColBuffer c)) doQuery uses the extractFns action to create the result buffer, and the two actions (extract and free) which are passed to doQuery2. > doQuery2 :: ErrorHandle -> StmtHandle -> IO a -> IO () -> [a] -> IO [a] > doQuery2 err stmt extractData freeMem results = do > rc <- fetch err stmt > if rc == oci_NO_DATA > then do > freeMem > return results > else do > v <- extractData > doQuery2 err stmt extractData freeMem (v:results) > doQuery :: Session -> String -> (ErrorHandle -> StmtHandle -> IO (IO a, IO ())) -> IO [a] > doQuery (Sess env err con) qry extractFns = do > stmt <- getStmt env err > prepare err stmt qry > execute err con stmt > (extractData, freeMem) <- extractFns err stmt > doQuery2 err stmt extractData freeMem [] The interface provided by doQuery means I have to write extract functions like this. Here's one for a select that returns three String columns. It's quite awkward... > ex3 :: ErrorHandle -> StmtHandle -> IO (IO (String, String, String), IO ()) > ex3 err stmt = do > (fetchcol1, freecol1) <- getExtractFnString 1 err stmt > (fetchcol2, freecol2) <- getExtractFnString 2 err stmt > (fetchcol3, freecol3) <- getExtractFnString 3 err stmt > return > ( do { s1 <- fetchcol1; s2 <- fetchcol2; s3 <- fetchcol3; return (s1, s2, s3) } > , do { freecol1; freecol2; freecol3 } > ) ***************************************************************** The information in this email and in any attachments is confidential and intended solely for the attention and use of the named addressee(s). This information may be subject to legal professional or other privilege or may otherwise be protected by work product immunity or other legal rules. It must not be disclosed to any person without our authority. If you are not the intended recipient, or a person responsible for delivering it to the intended recipient, you are not authorised to and must not disclose, copy, distribute, or retain this message or any part of it. ***************************************************************** From Tom.Pledger at peace.com Wed Sep 24 13:08:15 2003 From: Tom.Pledger at peace.com (Tom Pledger) Date: Tue Sep 23 20:08:28 2003 Subject: Database interface - would like advice on oracle library binding In-Reply-To: <7DFF3BC6CA957441AEADA7F340BFAA3401B33742@GBLONEX11.lon.invesco.com> References: <7DFF3BC6CA957441AEADA7F340BFAA3401B33742@GBLONEX11.lon.invesco.com> Message-ID: <16240.57455.812576.183573@tux-17.corp.peace.com> Bayley, Alistair writes: : | Still making slow progress on an Oracle database binding... now I'm trying | to fit the API I have into some sort of abstract interface (like the one(s) | discussed previously: | http://haskell.org/pipermail/haskell-cafe/2003-August/004957.html ). | | | 1. Is the left-fold the best/only interface to expose? I think yes, | but that doesn't allow fine-grained control over cursors i.e. being | able to open many cursors at once and interleave fetches from | them. Or does it? It looks like the interleaving would be limited to a nested loop structure: a cursor could be processed in full during one extraction for another cursor. Application-side nested loop structures are often a poor substitute for server-side joins. | 2. I'm finding it hard to write a doQuery function that takes an | extraction function that isn't a pig to write. Some advice would be | useful here... (and a long-ish explanation follows): : Here's my attempt to summarise the piggishness you describe: The interface to Oracle requires that you initialise a cursor by allocating a suitably typed buffer for each column prior to fetching the first row, and finalise a cursor by freeing those buffers after fetching the last row. This means that we must iterate over the columns 3 times. We would prefer to express this iteration only once, and have the other 2 happen automatically within the library. (As opposed to what ex3 does, which is to iterate for getExtractFnString, iterate for fetchcolN, and iterate for freecolN.) Here's one approach: find the OCI equivalent of JDBC's ResultSetMetaData, and use it to drive the allocation and freeing of buffers. Here's another: Add a mode attribute to the abstract data type which encompasses ErrorHandle and StmtHandle. (I'll persist in referring to that ADT as Cursor.) Expect extraction functions to be written along these lines: \cursor result -> do field1 <- getInt cursor field2 <- getString cursor field3 <- getString cursor return ((field1, field2, field3):result, True) Make getInt (and friends) behave differently depending on the mode of the cursor they're passed: either allocate a buffer and return _|_, decode and return the current column of the current row, or free a buffer and return _|_. doQuery could then apply the extraction function once in Allocate mode after opening the cursor, once per fetched row in Decode mode, and once in Free mode at the end. There's nothing to stop an extraction function from varying the number of get___ functions it applies, or trying to match their results when not in Decode mode. These weakness could be mitigated by: Pointing out that some database connection standards (JDBC, and for all I know also ODBC) don't guarantee that you can still get at a row's 1st column after you've looked at its 2nd column, i.e. there's a precedent for such fragility. Complicating the extraction functions by giving them the type (Cursor -> b -> IO (IO (b, Bool))) , expecting that all the get___ functions are applied in the outer IO layer, and undertaking that the inner IO layer will only be used in Decode mode. Regards, Tom From oleg at pobox.com Tue Sep 23 22:51:08 2003 From: oleg at pobox.com (oleg@pobox.com) Date: Wed Sep 24 00:51:28 2003 Subject: Database interface - would like advice on oracle library binding In-Reply-To: <7DFF3BC6CA957441AEADA7F340BFAA3401B33742@GBLONEX11.lon.invesco.com> Message-ID: <200309240451.h8O4p8ic023176@adric.fnmoc.navy.mil> > 1. Is the left-fold the best/only interface to expose? I think yes, but that > doesn't allow fine-grained control over cursors i.e. being able to open many > cursors at once and interleave fetches from them. Or does it? I'd like to remark first that many real databases let us avoid opening many cursors at once. It seems to be more efficient to do as much as possible within the engine -- using the full power of SQL or Stored Procedures -- and relay only the final results through the boundary. We can use SELECT ... UNION, correlated subqueries, SELECT rowid, self (outer) joins, etc. There are cases where SQL is powerless -- a transitive closure comes to mind. Stored procedures may sometimes help. A stored procedure may return a 'row' and thus can be used as a "generator" in Icon or Ruby's sense of the word. Still, opening several cursors may be unavoidable. The left fold approach may still help -- we _can_ mechanically invert a left fold combinator (to be precise, a half-baked version of it) into a lazy list. Please see a separate message "how to invert the left fold" > This approach is quite awkward (especially w.r.t. writing extract > functions), and it's hard for me to see how to build a better interface. > Hard, because of the memory management requirements. I believe the extract functions can be constructed automatically -- similar to the way Quickcheck constructs test cases. I believe that instead of > results <- doQuery dbconn sqltext [] \row results -> do > name <- stringv row 1 > address <- stringv row 2 > return (name,address):results we can do results <- doQuery dbconn sqltext [] process where process:: String -> String -> [(String,String)] -> [(String,String)] process name address results = (name,address):results doQuery should be able to figure our that the function 'process' takes two arguments of type string. Therefore, doQuery needs to extract two columns from the current row, both of type string. Note, 'process' is a pure function. There is no reason for it to be monadic (if it doesn't have to be). Because doQuery is a left-fold iterator, it can keep reusing buffers until the iteration finishes. We can keep the pointers to column buffers in a polytipic list, and we can keep the corresponding deallocation actions in a list of IO (). When we prepare a statement, we create both lists. When we finish doQuery, we scan the list of destructor actions and execute them to free all the buffers. It's hard for me to write the corresponding code because I don't have Oracle handy (besides, I like Informix better). Is it possible to come up with a "stub" that uses flat files? We are only interested in fetching rows. It doesn't matter if these rows come from a file or from a database. That would make prototyping the interface quite easier. From Alistair_Bayley at ldn.invesco.com Wed Sep 24 11:44:16 2003 From: Alistair_Bayley at ldn.invesco.com (Bayley, Alistair) Date: Wed Sep 24 05:47:58 2003 Subject: Database interface - would like advice on oracle library bind ing Message-ID: <7DFF3BC6CA957441AEADA7F340BFAA3401B33749@GBLONEX11.lon.invesco.com> > From: oleg@pobox.com [mailto:oleg@pobox.com] > > I'd like to remark first that many real databases let us avoid opening > many cursors at once. It seems to be more efficient to do as much as > possible within the engine ... I agree (and Tom Pledger makes the same comment). It's rare that I want to have more than one cursor open at once. When I've seen other developers do it, they're often doing something that would be better accomplished as a join. And that last time I remember doing it, it was a nested loops situation (which, on reflection, probably could have been done as a join... oh well). OTOH, I have seen interleaved fetches used to simulate a join where the specific DBMS product struggled to perform the join efficiently itself. In this case the developer is making up for a deficiency in the DBMS, and I don't want to exclude the option from the interface (if you agree with what you read on http://www.dbdebunk.com , then most DBMS products are deficient in a number of ways). > Still, opening several cursors may be unavoidable. The left fold > approach may still help -- we _can_ mechanically invert a left fold > combinator (to be precise, a half-baked version of it) into a lazy > list. Please see a separate message "how to invert the left fold" Found it. I'll have a look... > I believe the extract functions can be constructed automatically -- > similar to the way Quickcheck constructs test cases. I believe that > instead of I though this might be possible, but I had no idea how to do it. I'll have a look at the Quickcheck source to see how it's done, unless you can suggest a better example. > buffers. It's hard for me to write the corresponding code because I > don't have Oracle handy (besides, I like Informix better). Is it > possible to come up with a "stub" that uses flat files? We are only > interested in fetching rows. It doesn't matter if these rows come from > a file or from a database. That would make prototyping the interface > quite easier. I wasn't interested so much in prototyping the interface, as trying to write an implementation that supported the interface(s) discussed previously. I intended to provide the left-fold interface, and was wondering if that was all that was needed (for selects). Still, this would be a good exercise for me, at least so I can figure out how to generate extraction functions. > From: Tom Pledger [mailto:Tom.Pledger@peace.com] > Here's one approach: find the OCI equivalent of JDBC's > ResultSetMetaData, and use it to drive the allocation and freeing of > buffers. I've considered this, and I think it's the next route I'll take (the OCI supports it). At the least it'll give me the ability to perform arbitrary queries (at present I have to know the types of the result set columns and construct the extraction function manually). I've also considered stuffing more information into the Cursor type (which I've introduced), and using this in a modal fashion to decide what to do at each point. > Here's another: ... > Make getInt (and friends) behave differently depending on the mode > of the cursor they're passed: either allocate a buffer and return > _|_, decode and return the current column of the current row, or > free a buffer and return _|_. Not wanting to sound too dumb, but how do you return _|_ ? Thanks, Alistair. ***************************************************************** The information in this email and in any attachments is confidential and intended solely for the attention and use of the named addressee(s). This information may be subject to legal professional or other privilege or may otherwise be protected by work product immunity or other legal rules. It must not be disclosed to any person without our authority. If you are not the intended recipient, or a person responsible for delivering it to the intended recipient, you are not authorised to and must not disclose, copy, distribute, or retain this message or any part of it. ***************************************************************** From Keith.Wansbrough at cl.cam.ac.uk Wed Sep 24 12:11:44 2003 From: Keith.Wansbrough at cl.cam.ac.uk (Keith Wansbrough) Date: Wed Sep 24 06:12:10 2003 Subject: Database interface - would like advice on oracle library bind ing In-Reply-To: Your message of "Wed, 24 Sep 2003 10:44:16 BST." <7DFF3BC6CA957441AEADA7F340BFAA3401B33749@GBLONEX11.lon.invesco.com> Message-ID: > > Here's another: ... > > Make getInt (and friends) behave differently depending on the mode > > of the cursor they're passed: either allocate a buffer and return > > _|_, decode and return the current column of the current row, or > > free a buffer and return _|_. > > Not wanting to sound too dumb, but how do you return _|_ ? Return undefined or better, error "Erroneously demanded bottom result of buffer allocation phase" Note that undefined :: a undefined = undefined error :: String -> a {- defined internally -} HTH. --KW 8-) -- Keith Wansbrough http://www.cl.cam.ac.uk/users/kw217/ University of Cambridge Computer Laboratory. From brandon at its.caltech.edu Wed Sep 24 15:22:18 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Wed Sep 24 17:22:22 2003 Subject: Modeling multiple inheritance Message-ID: 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 From brandon at its.caltech.edu Wed Sep 24 20:02:32 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Wed Sep 24 22:02:36 2003 Subject: Modeling multiple inheritance In-Reply-To: References:

Message-ID: On Thu, 25 Sep 2003 ozone@algorithm.com.au wrote: > On 25/09/2003, at 7:22 AM, Brandon Michael Moore wrote: > > > 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. > ... > > Hi Brandon, it looks like the way that you're modelling inheritance and > OO-style overloading is basically the same way that I did in my thesis: > > http://www.algorithm.com.au/mocha > > The actual implementation of the thesis will be up in CVS in ~24 hours, > I'm just waiting from an email back from the people I'm getting it > hosted with. > > If you want a quick run-down on how I did the OO-style overloading > without delving into the paper, let me know and I'll post a quick > summary. I've only skimmed your email, but I think that the problem > you're having with interfaces is solved with the way I'm modelling OO > overloading and class inheritance. Thanks. I think I could use the summary. I already found and skimmed your thesis, and I don't think it gives me exactly what I want. All you do in chapter 3 is represent a multiple inheritance hierarcy. I want default instances that will propagate method definitions along the hierarcy. I'm not sure that's possible though. I want something like this: data Object data ClassA data ClassB data ClassC class SubClass super sub instance SubClass Object ClassA instance SubClass Object ClassB instance SubClass ClassA ClassC instance SubClass ClassB ClassC class HasFooMethod cls args result foo :: cls -> args -> result instance SubClass super sub, HasFooMethod super args result ,??? => HasFooMethod sub args result where foo obj args = foo (upCast obj) args instance HasFooMethod Object int int where foo = id (now all four classes have a foo method) Brandon From simonpj at microsoft.com Thu Sep 25 09:19:26 2003 From: simonpj at microsoft.com (Simon Peyton-Jones) Date: Thu Sep 25 03:19:30 2003 Subject: Modeling multiple inheritance Message-ID: <4B93206CA3C55D4F96A61C9B1DC80DE350CA0B@EUR-MSG-03.europe.corp.microsoft.com> When Mark Shields and I tackled this problem we came up with Object-Oriented Style Overloading for Haskell http://research.microsoft.com/~simonpj/Papers/oo-haskell/index.htm It describes an (unimplemented) extension to Haskell, rather than modelling it in Haskell itself, but you may find it interesting none the less. Simon | -----Original Message----- | From: haskell-cafe-bounces@haskell.org [mailto:haskell-cafe-bounces@haskell.org] On Behalf Of | Brandon Michael Moore | Sent: 24 September 2003 22:22 | To: haskell-cafe@haskell.org | Subject: Modeling multiple inheritance | | 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 | | _______________________________________________ | Haskell-Cafe mailing list | Haskell-Cafe@haskell.org | http://www.haskell.org/mailman/listinfo/haskell-cafe From simons at cryp.to Tue Sep 23 17:26:42 2003 From: simons at cryp.to (Peter Simons) Date: Thu Sep 25 06:57:58 2003 Subject: Generating "setMember" functions for record structures In-Reply-To: References: <87wubzr6wk.fsf@peti.cryp.to> Message-ID: <87fzinr36l.fsf@peti.cryp.to> Peter Gammie writes: > Haskell Report, Sec 3.15.3 Great, that's exactly what I need. Thanks a lot to all who replied! Peter From ozone at algorithm.com.au Thu Sep 25 12:21:59 2003 From: ozone at algorithm.com.au (ozone@algorithm.com.au) Date: Thu Sep 25 06:57:59 2003 Subject: Modeling multiple inheritance In-Reply-To: References: Message-ID: On 25/09/2003, at 7:22 AM, Brandon Michael Moore wrote: > 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. ... Hi Brandon, it looks like the way that you're modelling inheritance and OO-style overloading is basically the same way that I did in my thesis: http://www.algorithm.com.au/mocha The actual implementation of the thesis will be up in CVS in ~24 hours, I'm just waiting from an email back from the people I'm getting it hosted with. If you want a quick run-down on how I did the OO-style overloading without delving into the paper, let me know and I'll post a quick summary. I've only skimmed your email, but I think that the problem you're having with interfaces is solved with the way I'm modelling OO overloading and class inheritance. -- % Andre Pang : trust.in.love.to.save From oleg at pobox.com Wed Sep 24 21:20:00 2003 From: oleg at pobox.com (oleg@pobox.com) Date: Thu Sep 25 06:58:00 2003 Subject: Database interface - would like advice on oracle library binding In-Reply-To: <7DFF3BC6CA957441AEADA7F340BFAA3401B33749@GBLONEX11.lon.invesco.com> Message-ID: <200309250320.h8P3K0mh025061@adric.fnmoc.navy.mil> The following code illustrates a _generic_ interface to low-level database code. The left-fold iterator doQuery is completely generic over any possible iterator -- no matter how many columns the query returns, what are the types of these columns and what is the type of the seed (accumulator). The code for doQuery remains the same. The iterator allocates buffers for columns at the beginning and frees the buffers at the very end. Again, this buffer handling is generic. There is no longer need to write extraction/unmarshalling function for specific types of rows. We only need fetching functions for specific datatypes (not columns!). Again, the query and the row buffer management code is completely generic. I guess I'm repeating myself. The tests: -- Query returns one column of type String -- Never mind undefined: we return some static data in the buffers, -- we don't have any oracle to bind to test1 = doQuery undefined undefined iter1 ([]::[String]) where iter1:: String -> [String] -> Either [String] [String] iter1 s acc = Right $ s:acc -- Query returns two columns of types String and Int test2 = doQuery undefined undefined iter2 ([]::[(String,Int)]) where iter2:: String -> Int -> [(String,Int)] -> Either [(String,Int)] [(String,Int)] iter2 s i acc = Right $ (s,i):acc -- Query returns three columns of types Int, String and Int test3 = doQuery undefined undefined iter3 ([]::[(Int,String,Int)]) where iter3:: Int -> String -> Int -> [(Int,String,Int)] -> Either [(Int,String,Int)] [(Int,String,Int)] iter3 i1 s i2 acc = Right $ (i1,s,i2):acc Use the function runtests to run either of these tests. The code follows. Compiler flags: -fglasgow-exts -fallow-overlapping-instances -- DB column buffers type BufferSize = Int data BufferType = ORA_char | ORA_int type Position = Int -- column number of the result table data Buffer = Buffer { bufptr :: String -- for this stub, just use String , nullindptr :: String -- likewise , retsizeptr :: String -- likewise , size:: BufferSize , pos:: Position , ora_type:: BufferType } -- understandably, below is just a stub ... alloc_buffer (siz, typ) ps = return $ Buffer { bufptr = show ps, pos = ps, size = siz, ora_type = typ} -- In this stub, don't do anything free ptr = return () -- DB Column types class DBType a where alloc_buffer_hints:: a -> (BufferSize, BufferType) col_fetch:: Buffer -> IO a instance DBType String where alloc_buffer_hints _ = (2000, ORA_char) col_fetch buffer = return (bufptr buffer) instance DBType Int where alloc_buffer_hints _ = (4, ORA_int) col_fetch buffer = return (read $ bufptr buffer) -- need to add more ... -- Row iteratees. Note, the folowing two instances cover ALL possible -- iteratees. No other instances are needed class SQLIteratee iter seed where iter_apply:: [Buffer] -> seed -> iter -> IO (Either seed seed) alloc_buffers:: Position -> iter -> seed -> IO [Buffer] instance (DBType a) => SQLIteratee (a->seed->Either seed seed) seed where iter_apply [buf] seed fn = col_fetch buf >>= (\v -> return$ fn v seed) alloc_buffers n _ _ = sequence [alloc_buffer (alloc_buffer_hints (undefined::a)) n] instance (SQLIteratee iter' seed, DBType a) => SQLIteratee (a->iter') seed where iter_apply (buf:others) seed fn = col_fetch buf >>= (\v -> iter_apply others seed (fn v)) alloc_buffers n fn seed = do this_buffer <- alloc_buffer (alloc_buffer_hints (undefined::a)) n other_buffers <- alloc_buffers (n+1) (fn (undefined::a)) seed return (this_buffer:other_buffers) free_buffers = mapM_ free -- The left fold iterator -- the query executor data Session -- not relevant for this example data SQLStmt db_execute session query = return () db_fetch_row buffers = return () -- use static data doQuery:: (SQLIteratee iter seed) => Session -> SQLStmt -> iter -> seed -> IO seed -- In this example, we just allocate buffers, "fetch" two rows and terminate -- with a clean-up doQuery session query iteratee seed = do buffers <- alloc_buffers 0 iteratee seed db_execute session query db_fetch_row buffers (Right seed) <- iter_apply buffers seed iteratee db_fetch_row buffers (Right seed) <- iter_apply buffers seed iteratee free_buffers buffers return seed -- Tests -- Query returns one column of type String test1 = doQuery undefined undefined iter1 ([]::[String]) where iter1:: String -> [String] -> Either [String] [String] iter1 s acc = Right $ s:acc -- Query returns two columns of types String and Int test2 = doQuery undefined undefined iter2 ([]::[(String,Int)]) where iter2:: String -> Int -> [(String,Int)] -> Either [(String,Int)] [(String,Int)] iter2 s i acc = Right $ (s,i):acc -- Query returns three columns of types Int, String and Int test3 = doQuery undefined undefined iter3 ([]::[(Int,String,Int)]) where iter3:: Int -> String -> Int -> [(Int,String,Int)] -> Either [(Int,String,Int)] [(Int,String,Int)] iter3 i1 s i2 acc = Right $ (i1,s,i2):acc runtests test = test >>= (mapM_ $ putStrLn . show) From list at taesch.com Thu Sep 25 16:55:29 2003 From: list at taesch.com (Luc Taesch) Date: Thu Sep 25 08:55:32 2003 Subject: Haskell as specification language Message-ID: <200309251255.h8PCtO93011368@fury.vortech.net> out of curiosity, is haskell already been used as a "specification" language ? i was thinking in a business term, rather than mathematical one. (i.e. one than " normal mortal can read, even with a bit of training ;-) I.e. one would specifiy a model, an application ( possibly run it on samples), and hand it over to developper, typically in other languages/environment... I am aware of the contracts paper of SPJ, for instance... others you may think of ? does a spec. always leads to a DSL , for instance ? I imagine a specification for an application resulting in a DSL for the "data model"/ " process model " part, (which is the generic/ reusabel part, IMH) plus something more specific to the effective application targeted... (I cannot resist explaining my long term view : if IT outsourcing really takes off one day, one key factor will be proper specification, which is in a dear state in the current practice nowadays "in the field") ref: http://search.ft.com/search/article.html?id=030923007625 From brandon at its.caltech.edu Thu Sep 25 14:25:26 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Thu Sep 25 16:25:30 2003 Subject: lexer puzzle In-Reply-To: <001101c38388$85f3e390$1e02a8c0@Sean> References: <9D83635E-E77C-11D7-9A3E-000393D3FD08@cs.uu.nl><200309240044.32206.qrczak@knm.org.pl> <3F722BEA.2090805@cse.ogi.edu> <001101c38388$85f3e390$1e02a8c0@Sean> Message-ID: Note I've replied to haskell-cafe. This post is a bit chatty and low on solid answers. On Thu, 25 Sep 2003, Sean L. Palmer wrote: > > >>A... should be split into "A.." and "." > > >I found a compromise: let's make it a lexing error! :-) > > At least that agrees with what some Haskell compilers implement. No > > current Haskell compiler/interpreter agrees with what the report seems > > to say, that is that "A..." should be lexed as the two tokens "A.." and > > ".", and similarly, "A.where" should be lexed as "A.wher" followed by "e". > > Hi. I'm really new to Haskell, just learning it, and I must say I'm pretty > overwhelmed by the large variety of constructs. (=>, <-, \ to name a few) Would that be \ as in TREX row variable polymorphism? Just remember most operators are just library functions. It's only =, ->, =>, <-, :: that are really part of the language, and {,},; for grouping. Did I miss any? > > But I'm just writing this to let you guys know (surely you know this > already) that anyone from a C/C++/Java/Delphi background is going to > completely misunderstand the meaning of A.anything in Haskell... it's > completely nonintuitive to people with my background. I kinda like dot > notation because it ties together the symbols visually, for instance > "myrec.myfield" is more of a unit than "myrec myfield". It stays together > better when surrounded by other code, and would result in fewer parenthesis > necessary. A Python programmer would understand instantly: Python uses exactly the same syntax for module access, though Python modules are usually in lowercase. It also seems to be very much in the spirit of "access a member of this object" of an OO language. Or was that supposed to be composition of a constructor with a function, A . f? Function composition, and higher order functions in general are likely to confuse an imperative programmer, but I think there isn't much syntax can do there. Or are you talking about the field access syntax? Maybe the problem is that dot has two to five different meanings, function composition, naming module members, building hierarchial module names, being a decimal point, and making elipses, and is commonly used for yet another purpose in OO languages. > Haskell to me seems to be a great language with a syntax problem, and a bad > case of too many ways to do the same thing; thus every programmer does > things their own way and it's difficult to grasp the language by looking at > various programs, since they're all so very different. As a small example, > there's 'let' vs. 'where'. Maybe a bit of pruning would be in order. Do you mean the syntax is bad in places? Haskell is the cleanest language I know of, but I'm sure it has some grungy bits. I've had problems with unary minus (can't slice binary minus), and precedence of with irrefuatble patterns and type ascription. I would be happy for any confusing syntax to be improved. Any good ideas? Syntax change is a possibility: do notation is a relatively recent addition, and arrow syntax is in the works. I think you might instead mean the syntax cuts down our market share because it isn't like common (C derived) languages. I don't think Haskell could borrow any more syntax from C without actually making the language worse. It's a problem, but not with the syntax. If someone is so solidly into a C++/Java/OO mindset that the syntax would be a problem, the semantics would probably be even more of a problem. I would suggest Python if Haskell was too much of a jump for someone. It's still OO, but it encourages more flexible and interesting programs, and you don't have to live in a Java type system. Plus, it has more libraries, bindings, and PR, so it's easier to get permission to use it in a company. If someone is used to Python's layout rule and lack of type signatures, and gets their head around some of the fun you can have dynamically picking which members of an object to access, assigning to __dict__ and so on, then Haskell should be much less of a jump in syntax, and less imposing in semantics. > That said, I still think it looks more promising than any other language > I've looked at that actually is being actively used and maintained and has a > decent installed base and good cross platform support. So I will learn it. > I just wish the transition was easier and that it took less time to learn. > ;) > > Sean I learned Haskell from the "gentle introduction". It seemed gentle enough to me but others disagree, so I'm probably not the best for advice for the raw beginner. If you are interested in learning about monads though, Jeff Newbern's monad tutorial seems accessible and as complete as anything this side of Phil Wadler's paper. I hope learning Haskell goes well. Brandon From oleg at pobox.com Thu Sep 25 20:20:02 2003 From: oleg at pobox.com (oleg@pobox.com) Date: Thu Sep 25 22:20:24 2003 Subject: Modeling multiple inheritance In-Reply-To: Message-ID: <200309260220.h8Q2K2vZ026705@adric.fnmoc.navy.mil> Brandon Michael Moore wrote: > 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 > One problem is that the subclass relationship needs the functional > dependency > 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? Yes. The code included. The solution is trivial: in case of a multiple inheritance, a class has a _sequence_ of superclasses rather than a single superclass. Like instance SubClass (Object,()) ClassA instance SubClass (Object,()) ClassB -- Multiple inheritance (including the diamond!) instance SubClass (ClassA,(ClassB,())) ClassC instance SubClass (ClassA,(ClassB,(ClassC,()))) ClassD And we need some intelligence to traverse the sequence. But even a computer can do that. I would like to propose a different solution: a dual of typeclasses in the value domain. Function foo is just a regular function foo:: Object -> Int -> Int foo x y = y We then need a class MApplicable fn args result with a method mapply. The trick is that the method should take any object of a type castable and cast it to the type of the first argument of fn. The cast can be made safe and statically checkable, using the type heap. Actually, we can use the type heap to model the dispatch table (whose rows are functions and columns are object/classes). Given a function and an object, we can search in many way for the applicable combination. And now, the code for the solution that works. Compiler flags: -fglasgow-exts -fallow-overlapping-instances -fallow-undecidable-instances data Object = Object data ClassA = ClassA data ClassB = ClassB data ClassC = ClassC data ClassD = ClassD class SubClass super sub | sub -> super where upCast :: sub -> super instance SubClass (Object,()) ClassA instance SubClass (Object,()) ClassB -- Multiple inheritance (including the diamond!) instance SubClass (ClassA,(ClassB,())) ClassC instance SubClass (ClassA,(ClassB,(ClassC,()))) ClassD class HasFooMethod cls args result where foo :: cls -> args -> result instance (SubClass supers sub, HasFooMethod supers args result) => HasFooMethod sub args result where foo obj args = foo (upCast obj) args instance (HasFooMethod cls args result) => HasFooMethod (cls,()) args result where foo (x,()) = foo x instance (HasFooMethod cls args result) => HasFooMethod (x,cls) args result where foo (x,y) = foo y instance HasFooMethod Object Int Int where foo _ x = x test1::Int = foo Object (1::Int) test2::Int = foo ClassA (2::Int) test3::Int = foo ClassD (3::Int) -- Likewise for another method: class HasBarMethod cls args result where bar :: cls -> args -> result instance (SubClass supers sub, HasBarMethod supers args result) => HasBarMethod sub args result where bar obj args = bar (upCast obj) args instance (HasBarMethod cls args result) => HasBarMethod (cls,()) args result where bar (x,()) = bar x instance (HasBarMethod cls args result) => HasBarMethod (x,cls) args result where bar (x,y) = bar y instance HasBarMethod ClassB Bool Bool where bar _ x = x test4::Bool = bar ClassB True test5::Bool = bar ClassC True test6::Bool = bar ClassD True From ketil at ii.uib.no Fri Sep 26 09:59:12 2003 From: ketil at ii.uib.no (Ketil Z. Malde) Date: Fri Sep 26 02:59:17 2003 Subject: lexer puzzle In-Reply-To: References: <9D83635E-E77C-11D7-9A3E-000393D3FD08@cs.uu.nl> <200309240044.32206.qrczak@knm.org.pl> <3F722BEA.2090805@cse.ogi.edu> <001101c38388$85f3e390$1e02a8c0@Sean> Message-ID: Brandon Michael Moore writes: > Or was that supposed to be composition of a constructor with a function, A > . f? Function composition, and higher order functions in general are > likely to confuse an imperative programmer, but I think there isn't much > syntax can do there. I think there is a problem with too much overloaded syntax. Perhaps it is time to put non-ASCII characters to good use? For instance, function composition could use the degree sign: ? and leave the . for module qualification. Template Haskell could use double-angle quotation marks: ? ? and the section sign: ? and avoid clashing with list comprehensions and the function application operator. Implicit parameters could use an inverted question mark: ? And so on, just look for places where the semantics depend on spaces in the right (or wrong) place. -kzm -- If I haven't seen further, it is by standing in the footprints of giants From john at repetae.net Fri Sep 26 01:16:55 2003 From: john at repetae.net (John Meacham) Date: Fri Sep 26 03:16:59 2003 Subject: lexer puzzle In-Reply-To: References: <9D83635E-E77C-11D7-9A3E-000393D3FD08@cs.uu.nl> <200309240044.32206.qrczak@knm.org.pl> <3F722BEA.2090805@cse.ogi.edu> <001101c38388$85f3e390$1e02a8c0@Sean> Message-ID: <20030926071655.GA3520@momenergy.repetae.net> On Fri, Sep 26, 2003 at 08:59:12AM +0200, Ketil Z. Malde wrote: > Brandon Michael Moore writes: > > > Or was that supposed to be composition of a constructor with a function, A > > . f? Function composition, and higher order functions in general are > > likely to confuse an imperative programmer, but I think there isn't much > > syntax can do there. > > I think there is a problem with too much overloaded syntax. Perhaps > it is time to put non-ASCII characters to good use? > > For instance, function composition could use the degree sign: ? > and leave the . for module qualification. why not the actual functional composition operator: ? or ? we could also make good use of ? ? ? ? ? ? and all the other fun mathematical operators. > > Template Haskell could use double-angle quotation marks: ? ? > and the section sign: ? > and avoid clashing with list comprehensions and the function > application operator. > > Implicit parameters could use an inverted question mark: ? > > And so on, just look for places where the semantics depend on spaces > in the right (or wrong) place. I would love to be able to use unicode to make my programs more readable. just as an alternate syntax for awkward ascii constructs. and as operator, function names when they make sense. this could probably be done with a preprocessor, but wolud be easier in the compiler to work out the layout rule and handle language extensions and whatnot. John -- --------------------------------------------------------------------------- John Meacham - California Institute of Technology, Alum. - john@foo.net --------------------------------------------------------------------------- From ketil at ii.uib.no Fri Sep 26 10:56:40 2003 From: ketil at ii.uib.no (Ketil Z. Malde) Date: Fri Sep 26 03:56:50 2003 Subject: lexer puzzle In-Reply-To: <20030926071655.GA3520@momenergy.repetae.net> References: <9D83635E-E77C-11D7-9A3E-000393D3FD08@cs.uu.nl> <200309240044.32206.qrczak@knm.org.pl> <3F722BEA.2090805@cse.ogi.edu> <001101c38388$85f3e390$1e02a8c0@Sean> <20030926071655.GA3520@momenergy.repetae.net> Message-ID: John Meacham writes: >> For instance, function composition could use the degree sign: ? >> and leave the . for module qualification. > why not the actual functional composition operator: ? or ? Because: a) I've always used a small circle, the centered dot is for (dot) products. I guess this is just a matter of mathematical dialects. And b) I didn't find it :-) -------------- next part -------------- > we could also make good use of $B"O(B $B"P(B $B"M(B $B"+(B $B"J(B $B"K(B and all the other fun > mathematical operators. Cool! However, I think most/some current tools use ISO-8859(-1 or whatever) input, and for legacy reasons it may be a good idea to stick to symbols in that (those) subset(s). As you may have noticed, I suggested mostly these symbols for the language extensions, keeping H98 in 7 bits may or may not be a priority. At any rate, extensions could probably more easily disregrard legacy. > I would love to be able to use unicode to make my programs more > readable. just as an alternate syntax for awkward ascii constructs. > and as operator, function names when they make sense. Another thing; it should be possible to have (X)Emacs use display the glyphs you mention ($B"O(B $B"P(B $B"M(B $B"+(B $B"J(B $B"K(B) instead of the underlying multigraphs. > this could probably be done with a preprocessor, but wolud be easier in > the compiler to work out the layout rule and handle language extensions > and whatnot. Layout may be a problem. Not for type signatures, though. -kzm -- If I haven't seen further, it is by standing in the footprints of giants From Keith.Wansbrough at cl.cam.ac.uk Fri Sep 26 12:08:27 2003 From: Keith.Wansbrough at cl.cam.ac.uk (Keith Wansbrough) Date: Fri Sep 26 06:08:34 2003 Subject: lexer puzzle In-Reply-To: Your message of "Thu, 25 Sep 2003 13:25:26 PDT." Message-ID: > > Hi. I'm really new to Haskell, just learning it, and I must say I'm pretty > > overwhelmed by the large variety of constructs. (=>, <-, \ to name a few) > > Would that be \ as in TREX row variable polymorphism? Just remember most > operators are just library functions. It's only =, ->, =>, <-, :: that are > really part of the language, and {,},; for grouping. Did I miss any? Yes, you missed \, which is used for function abstraction: (\x -> x*x) 3 And ( , ) for tuples. And I don't think "->" is part of the language - it only appears in the type syntax, not term syntax. If you allow it, you have to allow * as well. --KW 8-) -- Keith Wansbrough http://www.cl.cam.ac.uk/users/kw217/ University of Cambridge Computer Laboratory. From Malcolm.Wallace at cs.york.ac.uk Fri Sep 26 12:12:58 2003 From: Malcolm.Wallace at cs.york.ac.uk (Malcolm Wallace) Date: Fri Sep 26 06:15:12 2003 Subject: lexer puzzle In-Reply-To: References:

Message-ID: <20030926111258.6bbde80f.Malcolm.Wallace@cs.york.ac.uk> Keith Wansbrough writes: > > And I don't think "->" is part of the language - it only appears in the type > syntax, not term syntax. If you allow it, you have to allow * as well. Errm, you just gave an example of -> in the term syntax... > (\x -> x*x) 3 Regards, Malcolm From Keith.Wansbrough at cl.cam.ac.uk Fri Sep 26 12:17:08 2003 From: Keith.Wansbrough at cl.cam.ac.uk (Keith Wansbrough) Date: Fri Sep 26 06:17:14 2003 Subject: lexer puzzle In-Reply-To: Your message of "Fri, 26 Sep 2003 11:12:58 BST." <20030926111258.6bbde80f.Malcolm.Wallace@cs.york.ac.uk> Message-ID: > Keith Wansbrough writes: > > > > And I don't think "->" is part of the language - it only appears in the type > > syntax, not term syntax. If you allow it, you have to allow * as well. > > Errm, you just gave an example of -> in the term syntax... > > > (\x -> x*x) 3 Guilty... sorry! :-( --KW 8-) -- Keith Wansbrough http://www.cl.cam.ac.uk/users/kw217/ University of Cambridge Computer Laboratory. From franka at cs.uu.nl Fri Sep 26 15:13:26 2003 From: franka at cs.uu.nl (Frank Atanassow) Date: Fri Sep 26 08:13:30 2003 Subject: lexer puzzle In-Reply-To: <20030926071655.GA3520@momenergy.repetae.net> Message-ID: On vrijdag, sep 26, 2003, at 09:16 Europe/Amsterdam, John Meacham wrote: > On Fri, Sep 26, 2003 at 08:59:12AM +0200, Ketil Z. Malde wrote: >> I think there is a problem with too much overloaded syntax. Perhaps >> it is time to put non-ASCII characters to good use? >> >> For instance, function composition could use the degree sign: ? >> and leave the . for module qualification. > > why not the actual functional composition operator: ? or ? > > we could also make good use of ? ? ? ? ? ? and all the other fun > mathematical operators. This is very readable, but not very writable. Until I get a keyboard with a ? key, I would prefer to restrict the syntax to ASCII/whatever and simply make it the editor's responsibility to display source code using special characters. ?????? From brandon at its.caltech.edu Fri Sep 26 13:57:04 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Fri Sep 26 15:57:10 2003 Subject: Modeling multiple inheritance In-Reply-To: <200309260220.h8Q2K2vZ026705@adric.fnmoc.navy.mil> References: <200309260220.h8Q2K2vZ026705@adric.fnmoc.navy.mil> Message-ID: On Thu, 25 Sep 2003 oleg@pobox.com wrote: > > Brandon Michael Moore wrote: > > > 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 > > > One problem is that the subclass relationship needs the functional > > dependency > > > 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? > > Yes. The code included. The solution is trivial: in case of a multiple > inheritance, a class has a _sequence_ of superclasses rather than a > single superclass. Like > > instance SubClass (Object,()) ClassA > instance SubClass (Object,()) ClassB > > -- Multiple inheritance (including the diamond!) > instance SubClass (ClassA,(ClassB,())) ClassC > instance SubClass (ClassA,(ClassB,(ClassC,()))) ClassD > > And we need some intelligence to traverse the sequence. But even a > computer can do that. That should solve my problem. Putting all the superclasses in a tuple should work. I'm worried about large class hierarchies. If it works on the java.* classes I should be fine. Have you used this approach before? I'm worried about compile time, runtime costs from the casts (hopefully they compile out), and maybe exceeding maximum stack depth in context reduction. This is a clever solution. I like it. Now, is anyone up to encoding the Dylan MRO in Haskell type classes? ;) > I would like to propose a different solution: a dual of > typeclasses in the value domain. Function foo is just a regular > function > > foo:: Object -> Int -> Int > foo x y = y > > We then need a class MApplicable fn args result with a method > mapply. The trick is that the method should take any object of a type > castable and cast it to the type of the first argument of fn. The cast > can be made safe and statically checkable, using the type > heap. Actually, we can use the type heap to model the dispatch table > (whose rows are functions and columns are object/classes). Given a > function and an object, we can search in many way for the applicable > combination. What type heap? It sounds like you are talking about information from an OO runtime, or are you talking about the collection of instances. I tried a system where method names were also represented by data types, but without your solution for multiple inheritance I couldn't get the implementation inheritance I wanted. How would you implement this dispatch table? What are the advantages of this approach over the type class encoding? I'm worried that generating bindings would be a problem if the dispatch table needs to be a monolithic value with a very interesting type in some file. Brandon > And now, the code for the solution that works. > Compiler flags: > -fglasgow-exts -fallow-overlapping-instances -fallow-undecidable-instances > > data Object = Object > data ClassA = ClassA > data ClassB = ClassB > data ClassC = ClassC > data ClassD = ClassD > > class SubClass super sub | sub -> super where > upCast :: sub -> super > > instance SubClass (Object,()) ClassA > instance SubClass (Object,()) ClassB > -- Multiple inheritance (including the diamond!) > instance SubClass (ClassA,(ClassB,())) ClassC > instance SubClass (ClassA,(ClassB,(ClassC,()))) ClassD > > class HasFooMethod cls args result where > foo :: cls -> args -> result > > instance (SubClass supers sub, > HasFooMethod supers args result) > => HasFooMethod sub args result where > foo obj args = foo (upCast obj) args > > instance (HasFooMethod cls args result) => HasFooMethod (cls,()) args result > where > foo (x,()) = foo x > > instance (HasFooMethod cls args result) => HasFooMethod (x,cls) args result > where > foo (x,y) = foo y > > instance HasFooMethod Object Int Int where > foo _ x = x > > test1::Int = foo Object (1::Int) > test2::Int = foo ClassA (2::Int) > test3::Int = foo ClassD (3::Int) > > -- Likewise for another method: > > class HasBarMethod cls args result where > bar :: cls -> args -> result > > instance (SubClass supers sub, > HasBarMethod supers args result) > => HasBarMethod sub args result where > bar obj args = bar (upCast obj) args > > instance (HasBarMethod cls args result) => HasBarMethod (cls,()) args result > where > bar (x,()) = bar x > > instance (HasBarMethod cls args result) => HasBarMethod (x,cls) args result > where > bar (x,y) = bar y > > instance HasBarMethod ClassB Bool Bool where > bar _ x = x > > test4::Bool = bar ClassB True > test5::Bool = bar ClassC True > test6::Bool = bar ClassD True > > From oleg at pobox.com Fri Sep 26 21:19:29 2003 From: oleg at pobox.com (oleg@pobox.com) Date: Fri Sep 26 23:19:53 2003 Subject: Modeling multiple inheritance In-Reply-To: References: <200309260220.h8Q2K2vZ026705@adric.fnmoc.navy.mil> Message-ID: <200309270319.h8R3JTmT028485@adric.fnmoc.navy.mil> Brandon Michael Moore wrote regarding the first solution: chain of super-classes: > I'm worried about large class hierarchies. If it works on the > java.* classes I should be fine. Have you used this approach before? I'm > worried about compile time, runtime costs from the casts (hopefully they > compile out), and maybe exceeding maximum stack depth in context > reduction. I didn't use the approach for anything as complex as all java.* classes. The only run-time costs are evaluating the chain of fst . snd . fst . .... The length and the composition of the chain is statically known. Perhaps the compiler can do something smart here. The maximum length of the chain is the maximum depth of the inheritance tree. It shouldn't be too big. A cast from a subclass to a superclass has to be executed anyway (if not by your code then by JVM). If the maximum stack depth is exceeded, we can repeat the compilation with a compiler flag to allocate a bigger stack. In my experience the only time I've seen the derivation stack depth exceeded is when the derivation truly diverges. > What type heap? It sounds like you are talking about information from an > OO runtime, or are you talking about the collection of instances. The other solution I talked so confusingly before is that of generic functions. For that, we need a way to obtain a value representation of a type. Several such representations exists: e.g., Typable, representation as an integer, etc. All our objects must be members of the class Typable. A method (generic function foo) would have the following signature: foo:: (Typable object) => object -> Int ->Int For example, if foo is defined for ClassA object only, we can write foo obj arg = if inherit_from (typeof obj) (typeof (undefined::ClassA)) then particular_instance_foo (coerce obj) arg else error "miscast" If bar is defined for classB and redefined in classC, we can write bar obj arg = if inherit_from (typeof obj) (typeof (undefined::ClassC)) then particular_instance1_bar (coerce obj) arg else if inherit_from (typeof obj) (typeof (undefined::ClassB)) then particular_instance2_bar (coerce obj) arg else error "miscast" The functions inherit_from and coerce avail themselves of a table that records the relationship between types using their value representations. The disadvantage of this approach is that the cast errors become run-time errors. OTH, because type representations and the whole inheritance graph are values, we can do much more. We can check for proper and improper diamond inheritance, we can do a rather sophisticated dispatch. Types heap and several ways of doing safe casts are discussed in http://www.haskell.org/pipermail/haskell/2003-August/012372.html http://www.haskell.org/pipermail/haskell/2003-August/012355.html See also: http://citeseer.nj.nec.com/cheney02lightweight.html http://citeseer.nj.nec.com/context/1670116/0 The Sketch of a Polymorphic Symphony http://homepages.cwi.nl/~ralf/polymorphic-symphony/ From dominic.steinitz at blueyonder.co.uk Sat Sep 27 09:02:22 2003 From: dominic.steinitz at blueyonder.co.uk (Dominic Steinitz) Date: Sat Sep 27 03:05:11 2003 Subject: (Off-topic) Question about categories Message-ID: <000301c384c5$4c8defe0$1464a8c0@canterburysoftware.com> Graham, I'm not sure if anyone mentioned the examples of a poset and a monoid as categories. There is no "internal" structure in these. In the former, the objects are the elements and there is a morphism between a and b iff a <= b. A functor then becomes an order preserving map. In the latter, there is one object and the morphisms are the elements. The identity is the identity map and if x and y are two elements / morphisms then composition is xy. A functor is then a homomorphism. Dominic. From brandon at its.caltech.edu Sat Sep 27 03:25:38 2003 From: brandon at its.caltech.edu (Brandon Michael Moore) Date: Sat Sep 27 05:25:42 2003 Subject: Modeling multiple inheritance In-Reply-To: <200309270319.h8R3JTmT028485@adric.fnmoc.navy.mil> References: <200309260220.h8Q2K2vZ026705@adric.fnmoc.navy.mil> <200309270319.h8R3JTmT028485@adric.fnmoc.navy.mil> Message-ID: On Fri, 26 Sep 2003 oleg@pobox.com wrote: > > Brandon Michael Moore wrote regarding the first solution: chain of > super-classes: > > > I'm worried about large class hierarchies. If it works on the > > java.* classes I should be fine. Have you used this approach before? I'm > > worried about compile time, runtime costs from the casts (hopefully they > > compile out), and maybe exceeding maximum stack depth in context > > reduction. > > I didn't use the approach for anything as complex as all java.* > classes. The only run-time costs are evaluating the chain of fst . snd > . fst . .... I think I can use the pair types as phantom types on a reference type, so my casts will hopefully be the identity function. (.) should be small enough to inline, so GHC probably compiles id . id ... id to id. Correct? > The length and the composition of the chain is statically > known. Perhaps the compiler can do something smart here. The maximum > length of the chain is the maximum depth of the inheritance tree. It > shouldn't be too big. A cast from a subclass to a superclass has to be > executed anyway (if not by your code then by JVM). If the maximum > stack depth is exceeded, we can repeat the compilation with a compiler > flag to allocate a bigger stack. In my experience the only time I've > seen the derivation stack depth exceeded is when the derivation truly > diverges. Same for me, but I've never tried to model the java.* hierarchy either. I think you get a cast (fst in your code) for each parent of each ancestor along the inheritance path, which probably increses the count some. Your code doesn't quite work. The instances you gave only allow you to inherit from the rightmost parent. GHC's inference algorithm seems to pick one rule for a goal and try just that. To find instances in the first parent and in other parents it needs to try both. I think I'll just give up on inheriting methods, and generate unrelated instances for each class that needs one. Brandon From veger at cistron.nl Sun Sep 28 00:45:22 2003 From: veger at cistron.nl (Peter J. Veger) Date: Sat Sep 27 17:45:25 2003 Subject: (Off-topic) Question about categories Message-ID: <001c01c38540$a7551cc0$0a02a8c0@veger3> > I'm not sure if anyone mentioned the examples of a poset and a monoid as categories. > There is no "internal" structure in these. In the former, the objects are the > elements and there is a morphism between a and b iff a <= b. A functor then becomes > an order preserving map. In the latter, there is one object and the morphisms are > the elements. The identity is the identity map and if x and y are two elements / > morphisms then composition is xy. A functor is then a homomorphism. > Dominic. It is probably better to use the example of a preordered set. A Preordered set is a set with a relation <=, that is reflexive (x<=x) and transtive (x<=y & y<=z --> x<=z) (but not necessarily antisymmetric, x<=y & y<=x --> x=y as for a poset). If, in a preordered set, a is the unique morphism: x -> y and b the unique morphism: y -> x, then (writing x for the identity morphism id(x)) a.b=y and b.a=x Categorically, x and y are isomorphic. Concluding that x=y is not in the categorical language. Of course, you should distinguish the category of preorders (posets, monoids) from preorder (poset, monoid) as a category. As for literature, you have the classic: Saunders Mac Lane: Categories for the Working mathematician but the working computer scientist may prefer: Andrea Asperti and Guiseppe Longo: Categories, Types, and Structures (MIT Press, 1991) Michael Barr and Charles Wells: Category Theory for Computing Science (3rd ed., Les Publication CRM, 1999) idem, Introduction to Category Theory (www.let.uu.nl/esslli/Courses/barr-wells.html) Peter J. Veger, Best, Netherlands From mai99dnn at userv1.informatik.uni-leipzig.de Mon Sep 29 03:55:18 2003 From: mai99dnn at userv1.informatik.uni-leipzig.de (Patrick Scheibe) Date: Sun Sep 28 20:55:22 2003 Subject: haddock depencies to std-modules Message-ID: <200309290255.18215.mai99dnn@userv1.informatik.uni-leipzig.de> hello I use haddock to create documentation. Everything is fine except of the references to standard modules. If I am using standardtypes like Int or String haddock wants to create crossrefs to these modules and fails because of missing "sourcecode". Is there a simple way to fix this? I've heard of interface files to create right refs to existing docs, but that doesnt work. Whats the right way? Cheers Patrick From simonmar at microsoft.com Mon Sep 29 13:37:55 2003 From: simonmar at microsoft.com (Simon Marlow) Date: Mon Sep 29 07:43:24 2003 Subject: haddock depencies to std-modules Message-ID: <3429668D0E777A499EE74A7952C382D1E5A385@EUR-MSG-01.europe.corp.microsoft.com> > I use haddock to create documentation. Everything is fine > except of the > references to standard modules. If I am using standardtypes > like Int or > String haddock wants to create crossrefs to these modules and > fails because > of missing "sourcecode". > > Is there a simple way to fix this? I've heard of interface > files to create > right refs to existing docs, but that doesnt work. Whats the > right way? One way is to use the Haddock interface files that are distributed with GHC. I haven't put any effort into making the process easy yet - ideally Haddock should consult GHC's package database to find the interface files, but it doesn't currently do this. You'll find Haddock interface files in the same location as the documentation when you install GHC (eg. /usr/share/ghc-/html/base/base.haddock on a Linux system). Cheers, Simon From shlomif at vipe.technion.ac.il Mon Sep 29 16:20:11 2003 From: shlomif at vipe.technion.ac.il (Shlomi Fish) Date: Mon Sep 29 08:20:29 2003 Subject: RFC: "Haskell for Perl Programmers" Presentation Message-ID: At our local Perl Mongers club we have decided to hold a series of introductions to foreign language. Being a big fan of Haskell (except for any practical use), I wrote a "Haskell for Perl Programmers" intro: http://vipe.technion.ac.il/~shlomif/lecture/Perl/Haskell/slides/ It's more or less complete, but there may still be things to add or remove. Note that until now I did not write too much about the typing facilities of Haskell. I could, though. Regards, Shlomi Fish ---------------------------------------------------------------------- Shlomi Fish shlomif@vipe.technion.ac.il Home Page: http://t2.technion.ac.il/~shlomif/ An apple a day will keep a doctor away. Two apples a day will keep two doctors away. Falk Fish From timd at macquarie.com.au Mon Sep 29 23:43:17 2003 From: timd at macquarie.com.au (Tim Docker) Date: Mon Sep 29 08:43:30 2003 Subject: Database interface - would like advice on oracle library bind ing Message-ID: <675EC99F3F6BBF40BB989DC2E33E875D33F18C@nt_lon_exm01.macbank> Alistair Bayley wrote: > Still making slow progress on an Oracle database binding... > now I'm trying to fit the API I have into some sort of > abstract interface (like the one(s) discussed previously: Since the previous dicussion I've discovered Krasimir Angelov's HSQL interface as part of his HTOOLKIT: https://sourceforge.net/project/showfiles.php?group_id=65248 This has a uniform abstract interface similar to the one discussed, apparently implemented for MYSQL, ODBC, and PostgreSQL. I've only played with the MYSQL variant. If possible, it would seem sensible to provide a consistent interface for your Oracle binding. If not, it may be worth persuading him to generalise the interface. Then again, perhaps you knew about this already. Tim From Alistair_Bayley at ldn.invesco.com Mon Sep 29 15:18:20 2003 From: Alistair_Bayley at ldn.invesco.com (Bayley, Alistair) Date: Mon Sep 29 09:18:59 2003 Subject: Database interface - would like advice on oracle library bind ing Message-ID: <7DFF3BC6CA957441AEADA7F340BFAA3401B33763@GBLONEX11.lon.invesco.com> Yes, I know of it. Krasimir's code uses similar functions and types in the three bindings (MySQL, ODBC, PostgreSql), but the bindings do not share a common interface (AFAICT). Because they are so similar, making them support a common interface shouldn't be a lot of work. I was hoping the earlier discussions on database interfaces would converge to a solution most interested parties agreed on. I'm still sold on the left-fold interface discussed previously. Oleg claims it is sufficient, as it can be automatically converted to a stream-like interface. I bit of me agrees with you that I should be consistent with previous work, and another bit says make it simple and elegant, and this leads me towards Oleg's proposals (left-fold + auto-generated value extraction functions + auto-generated stream functions). Do others on this list (and elsewhere) have any further opinions as to how a database interface ought to look? > -----Original Message----- > From: Tim Docker [mailto:timd@macquarie.com.au] > Sent: 29 September 2003 13:43 > To: Bayley, Alistair; haskell-cafe@haskell.org > Subject: RE: Database interface - would like advice on oracle library > bind ing > > > > Alistair Bayley wrote: > > > Still making slow progress on an Oracle database binding... > > now I'm trying to fit the API I have into some sort of > > abstract interface (like the one(s) discussed previously: > > > Since the previous dicussion I've discovered Krasimir > Angelov's HSQL interface as part of his HTOOLKIT: > > https://sourceforge.net/project/showfiles.php?group_id=65248 > > This has a uniform abstract interface similar to the one > discussed, apparently implemented for MYSQL, ODBC, and > PostgreSQL. I've only played with the MYSQL variant. > > If possible, it would seem sensible to provide a consistent > interface for your Oracle binding. If not, it may be worth > persuading him to generalise the interface. > > Then again, perhaps you knew about this already. > > Tim ***************************************************************** The information in this email and in any attachments is confidential and intended solely for the attention and use of the named addressee(s). This information may be subject to legal professional or other privilege or may otherwise be protected by work product immunity or other legal rules. It must not be disclosed to any person without our authority. If you are not the intended recipient, or a person responsible for delivering it to the intended recipient, you are not authorised to and must not disclose, copy, distribute, or retain this message or any part of it. ***************************************************************** From droundy at abridgegame.org Tue Sep 30 08:26:14 2003 From: droundy at abridgegame.org (David Roundy) Date: Tue Sep 30 07:26:17 2003 Subject: swap leak problem Message-ID: <20030930112611.GA30484@jdj5.mit.edu> Hello. I've got a memory problem that is stumping me, and could use advice in trying to fix it. In brief, the heap of my application keeps increasing with time, until after a few hours it uses up all my swap and gets killed. There is no indication of a true memory leak. Heap profiling indicates that the haskell memory usage (including stacks, via -xt) remains below its initial peak value. Similarly, I've attached a sort of debugging counting finalizer to my foreignPtrs (a static int in C land), and it's clear that all the foreignPtrs I allocate are getting freed. Most of them are allocated with mallocForeignPtrBytes anyways, which should be about as safe an allocation method as one could use. The allocation pattern of my code is as follows. It reads a gzipped patch file into memory, parses it, then reads in a bunch of other files, and writes modified versions of them. Then it reads in the next patch and repeats. So it does large amounts of allocation and deallocation on a periodic basis. About half of the memory used is stored in foreignPtrs. With the exception of the patch files (which I'll get to later), all foreignPtrs are generated by mallocForeignPtrBytes or mallocForeignPtrArray. The patch files are gzip compressed, so I don't know their size until I've read them. I read the patch file one chunk at a time, and use reallocBytes to allocate a bit more memory each time. I don't expect in general to be calling reallocBytes more than four times per patch file. Of course, once I'm done I convert to a foreignPtr via newForeignPtr finalizerFree. My only idea as to the source of the problem is that it could be either a result of memory fragmentation or of not cleaning up before allocating. Neither theory is particularly convincing. Unfortunately, I'm at a loss as to how to track this problem down, so any advise would be appreciated. -- David Roundy http://www.abridgegame.org/darcs From droundy at abridgegame.org Tue Sep 30 08:31:03 2003 From: droundy at abridgegame.org (David Roundy) Date: Tue Sep 30 07:31:12 2003 Subject: swap leak problem In-Reply-To: <20030930112611.GA30484@jdj5.mit.edu> References: <20030930112611.GA30484@jdj5.mit.edu> Message-ID: <20030930113102.GB30484@jdj5.mit.edu> Oh yes, one more datum. If I run hs_perform_gc before each and every mallocForeignPtrBytes and reallocBytes, the leak goes much more slowly, and the code gets much further in the job before getting killed. But it still leaks, and still gets killed. -- David Roundy http://www.abridgegame.org