- Haskell-Cafe - Haskell.org

Re: ANN: yst 0.2.1
by spoon 04 Aug '09

04 Aug '09

This sounds great! I really like using yaml for web development. However, my last use case was for a non-techie, so I created a little desktop application for editing the "yamlbase", if you excuse the neologism. I think it worked quite well, because the user was then presented with a friendly looking system ( I even wrote a little calendar widget :) without having to spend the precious moneys on dynamic hosting. In the end, everyone was very happy. Actually, it was written in Ruby, and Shoes, and the data part was a lot less powerful than the yst query language there - but I think that an analogous tool would be useful; a GUI which creates yst scripts, would make yst a very powerful tool for non-techies. John

1 0

funct.prog. vs logic prog., practical Haskell
by Petr Pudlak 04 Aug '09

04 Aug '09

Hi all, I'd like to convince people at our university to pay more attention to functional languages, especially Haskell. Their arguments were that (1) Functional programming is more academic than practical. (2) They are using logic programming already (Prolog); why is Haskell better than Prolog (or generally a functional language better than a logic programming language)? (1) is easier to answer, there are a lots of applications at HaskellWiki, or elsewhere around the Internet, written in Haskell, OCaml, etc. Still, I welcome comments on your experience, for example, if you have written some larger-scale application in Haskell (or another a functional language) that is not at HaskellWiki, and perhaps if/why you would recommend doing so to other people. (2) is harder for me, since I've never programmed in Prolog or another language for logic programming. I'd be happy if anyone who is experienced in both Prolog and Haskell could elaborate the differences, pros & cons etc. Thanks, Petr

8 11

ANNOUNCE: The Haskell Platform 2009.2.0.2
by Don Stewart 04 Aug '09

04 Aug '09

We're pleased to announce the third release of the Haskell Platform: a single, standard Haskell distribution for everyone. The specification, along with installers (including Windows and Unix installers for a full Haskell environment) are available. Download the Haskell Platform 2009.2.0.2: http://hackage.haskell.org/platform/ The Haskell Platform is a single, standard Haskell distribution for every system, in the form of a blessed library and tool suite for Haskell distilled from the thousands of libraries on Hackage, along with installers for a wide variety of systems. It saves developers work picking and choosing the best Haskell libraries and tools to use for a task. When you install the Haskell Platform, you get the latest stable compiler, an expanded set of core libraries, additional development tools, and cabal-install – so you can download anything else you need from Hackage. What you get: http://hackage.haskell.org/platform/contents.html With regular time-based releases, we expect the platform will grow into a rich, indispensable development environment for all Haskell projects. Since the first release, there have been over 100,000 downloads of the installers. Thanks! -- The Platform Infrastructure Team

6 8

Re: Simple quirk in behavior of `mod`
by gladstein＠gladstein.com 04 Aug '09

04 Aug '09

7 7

[Haskell Cafe] Troubles with StateT and Parsec
by Paul Sujkov 03 Aug '09

03 Aug '09

Hi haskellers, I have a few problems using monad transformers. I have such two functions: parseSyslog :: StateT Integer Parser TimeStamp parseString :: StateT Integer Parser LogString and the following code: parseString = do -- string parse here, all in the form of lift $ <parser> stamp <- lift $ lexeme parseTimestamp -- <?> "timestamp" message <- lift $ manyTill anyToken eof -- <?> "message" return (LogString <...parsed values here...> (check stamp console message) <...more parsed values here...>) where check :: (Maybe TimeStamp) -> Console -> String -> Maybe TimeStamp check Nothing Syslog message = case (lift parse $ parseSyslog "" message) of Left err -> Nothing Right res -> Just res <...other clauses here...> this code seems quite intuitive to me, however it doesn't compile with a king error: Couldn't match kind `(* -> *) -> * -> *' against `?? -> ? -> *' When matching the kinds of `t :: (* -> *) -> * -> *' and `(->) :: ?? -> ? -> *' Probable cause: `lift' is applied to too many arguments In the first argument of `($)', namely `lift parse' I'm not so familiar with monad transformers whatsoever, so I'll be very happy if someone can show me the right way. The code compile nicely if I use "parse" line in a such way: check Nothing Syslog message = case (parse (evalStateT parseSyslog 0) "" message) of but this is not what I really want. To be accurate, here is the sequence which I do want to have in the code: some user state is initialized; parseString gets called many times and changes the state via call to the parseSyslog (that is the only function that really uses/affects user state, everything else is pure Parsec code with it's own internal state). Two main problems that I have now is: 1) impossibility to use parse/parseTest functions with the (StateT <state type> Parser <parse type>) argument. I want it to be lifted somehow, but cannot see how 2) too many lifts in the code. I have only one function that really affects state, but code is filled with lifts from StateT to underlying Parser Sorry if the questions are silly; any help is appreciated -- Regards, Paul Sujkov

4 4

Fwd: [Haskell-cafe] funct.prog. vs logic prog., practical Haskell
by Alberto G. Corona 03 Aug '09

03 Aug '09

The middle road could be Curry <http://en.wikipedia.org/wiki/Curry>, sorry, this Curry <http://www.informatik.uni-kiel.de/~curry/>, a functional-logic language. I know that curry has gained a lot of interest from prolog programmers. There are compilers from Curry to Prolog. It is a haskell 98 implementation (more or less) with all logic programming features. It can be seen also as a logic language with haskell syntax. Therefore, its syntax is more mathematical, rather that the ugly clause-based syntax of Prolog, that is at odds with anything except with pure aristothelian logic. 2009/8/2 Thomas ten Cate <ttencate(a)gmail.com> On Sun, Aug 2, 2009 at 12:25, Petr Pudlak<deb(a)pudlak.name> wrote: > > Hi all, > > > > I'd like to convince people at our university to pay more attention to > > functional languages, especially Haskell. Their arguments were that > > > > (1) Functional programming is more academic than practical. > > Which, even if it were true, is an argument *for* instead of *against* > teaching it at a university; that is what the word "academic" means, > after all... > > Thomas > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe >

1 0

ANN: yst 0.2.1
by John MacFarlane 03 Aug '09

03 Aug '09

I'm pleased to announce the release of yst, now available on HackageDB. yst generates static websites from YAML or CSV data files and StringTemplates. This approach combines the speed, security, and ease of deployment of a static website with the flexibility and maintainability of a dynamic site that separates presentation and data. The easiest way to get a feel for yst is to try it: cabal update cabal install yst yst create testsite cd testsite yst yst attempts to fill a niche between two kinds of site creation tools. On the one hand you have simple static site generators like webgen, webby, nanoc, and my old custom system using make and pandoc. On the other hand, you have dynamic web frameworks like rails and django. For my own smallish websites, I found that the dynamic frameworks were overkill. Nobody but me was going to edit the pages, and I didn't want the trouble of writing and deploying a dynamic site, setting up a web server, and administering a database. A static site would be faster, easier to deploy, and more secure. But the dynamic frameworks offered one thing that the static site generators did not: an easy way to separate data from presentation. This was becoming increasingly important to me as I found myself constantly updating the same information (say, publication data for a paper) in multiple places (say, a LaTeX CV and a differently formatted web listing of papers). What I wanted was a site generation tool that used YAML text files as a database and allowed different kinds of documents to be produced from the same data. I couldn't find anything that did just what I wanted, so I wrote yst. By way of illustration, here are the build instructions for HTML and LaTeX versions of a CV, plus a web page with a list of papers: - url: cv.html title: CV template: cv.st data_common: &cvdata contact: from contact.yaml jobsbyemployer: from jobs.yaml order by start group by employer degrees: from degrees.yaml order by year desc awards: from awards.yaml order by year desc group by title papers: from papers.yaml order by year desc where (not (type = 'review')) reviews: from papers.yaml order by year desc where type = 'review' talks: from talks.yaml where date < '2009-09-01' order by date desc group by title dissertations: from dissertations.yaml order by role then year group by role theses: from theses.yaml order by year then student courses: from courses.yaml order by number group by title data: <<: *cvdata html: yes - url: cv.tex title: CV inmenu: no template: cv.st layout: layout.tex.st data: <<: *cvdata html: yes - url: papers.html title: Papers template: papers.st data: papersbyyear: from papers.yaml order by year desc then title group by year yst's query language is limited, and there are lots of things you can do with a full-fledged database that you can't do with yst. But yst is ideal, I think, for small to medium data-driven sites that are maintained by a single person who likes working with plain text. It scratched my itch, anyway, and I release it in case anyone else has the same itch. Code, documentation, and bug reports: http://github.com/jgm/yst/tree/master John

5 7

Writing a pnm file
by CK Kashyap 03 Aug '09

03 Aug '09

Hi, Now that I've understood how to generate raster points of a line in Haskell - the next thing I want to do is generate a pnm file with it. I've done it in perl as of now. In perl, I can have a scalar variable $x contain a string of 256*256*3 bytes (for 24-bit 256x256 image) and set pixels using substr on LHS. I was wondering how I could do something similar in Haskell? sub setPixel{ my($x,$y,$red,$green,$blue)=@_; my$pixel=pack "CCC",$red,$green,$blue; my$offset=$WIDTH*$y*3 + $x*3; substr($image,$offset,3) = $pixel; } Regards, Kashyap

2 5

Hackage and version control
by Vasili I. Galchin 03 Aug '09

03 Aug '09

Hello, It seems to me that Hackage doesn't provide version control, e.g. check out and check in. Am I incorrect? Kind regards, Vasili

8 8

FastCGI error handling
by Victor Nazarov 03 Aug '09

03 Aug '09

I've been trying to write some simple web application in haskell using FastCGI, HDBC and HStringTemplate. I've got stuck with the following problem. HDBC throws some exceptions and I wanted them to be caught and logged. The code was: -- Main.hs module Main (main) where ... import Control.Concurrent import Network.FastCGI driver :: CGI CGIResult driver = ... main :: IO () main = runFastCGIConcurrent' forkIO 10 (handleErrors test) But all I've got on the page and in the log is "SomeException" string. I've tried the workaround: -- Main.hs module Main (main) where import Network.FastCGI import Control.Exception import System.IO driver :: CGI CGIResult driver = ... ... main :: IO () main = runFastCGI driver `catch` \ex -> hPutStr stderr $ show (ex::SomeException) And this worked fine. I've got a detaied SqlException string in the log file. So I've decided to update exception handling in CGI package and written the following module: -- Network/CGI/Monad/NewException.hs module Network.CGI.Monad.NewException where import Prelude hiding (catch) import Control.Exception import Control.Monad import Control.Monad.Writer import Control.Monad.Reader import Network.CGI.Monad hiding (throwCGI, catchCGI, tryCGI) import Network.CGI (outputInternalServerError) -- | Throw an exception in a CGI monad. The monad is required to be -- a 'MonadIO', so that we can use 'throwIO' to guarantee ordering. throwCGI :: (MonadCGI m, MonadIO m, Exception e) => e -> m a throwCGI = liftIO . throwIO -- | Catches any expection thrown by a CGI action, and uses the given -- exception handler if an exception is thrown. catchCGI :: (Exception e) => CGI a -> (e -> CGI a) -> CGI a catchCGI c h = tryCGI c >>= either h return handleCGI :: (Exception e) => (e -> CGI a) -> CGI a -> CGI a handleCGI = flip catchCGI -- | Catches any exception thrown by an CGI action, and returns either -- the exception, or if no exception was raised, the result of the action. tryCGI :: (Exception e) => CGI a -> CGI (Either e a) tryCGI (CGIT c) = CGIT (ReaderT (\r -> WriterT (f (runWriterT (runReaderT c r))))) where f = liftM (either (\ex -> (Left ex,mempty)) (\(a,w) -> (Right a,w))) . try handleErrors = handleCGI outputException outputException e = outputInternalServerError [show (e :: SomeException)] Then I've updated Main.hs accordingly: module Main (main) where ... import Control.Concurrent import Network.FastCGI hiding (throwCGI, catchCGI, tryCGI, handleErrors, outputException) import Network.CGI.Monad.NewException driver :: CGI CGIResult driver = ... main :: IO () -- main = runFastCGIConcurrent' forkIO 10 (handleErrors test) -- This allways gives internal error without a chance to find out what happens main = runFastCGI $ handleErrors test -- This works fine So one variant of main function works, the other allways gives no output. What can be the case? -- Victor Nazarov

1 1