October 2009 - Haskell-Cafe

Where do these warnings come from?
by Colin Paul Adams 04 Oct '09

04 Oct '09

Does anyone recognize which module/function would emit the following warnings? WARNING: nonstandard use of escape in a string literal LINE 11: '# Pandoc\r\n\r\nPandoc is a program for converting ... ^ HINT: Use the escape string syntax for escapes, e.g., E'\r\n'. -- Colin Adams Preston Lancashire

2 2

Need some help with Alex macros
by Niklas Broberg 04 Oct '09

04 Oct '09

Hi all, I'm going crazy trying to get Alex to do what I want. I have the following regexp macros: @octEscape = [0123]? $octdig{1,2} @hexEscape = 'u' $hexdig{4} @charEscape = '\\' (@octEscape | @hexEscape | b | t | n | f | r | \" | \') and the following rules: \' (. # [\'\\] | @charEscape) \' { \p s -> CharTok (readCharTok s) } \" (. # [\"\\] | @charEscape)* \" { \p s -> StringTok (readStringTok s) } I've tried a lot of variations on where to put the backslashes and quotes to get it to do what I want, what you see above is just one variation, but no matter what I try I can't get it to work. > alexScanTokens $ show '\n' *** Exception: lexical error Can anyone help shed some light on this for me? Why can't I get alex to recognize escapes inside character literals? Cheers, /Niklas

1 0

ANNOUNCE: SourceGraph 0.5.2.0
by Ivan Lazar Miljenovic 04 Oct '09

04 Oct '09

Yes, this is yet another SourceGraph announcement, this time for version 0.5.2.0 [1]. [1] http://hackage.haskell.org/package/SourceGraph-0.5.2.0 The changes in this release are as follows: * Shift overall analysis to the top and per-module analysis to the end, as suggested by Duncan Coutts. * Graph drawing fixups: instances are now drawn correctly, and the module graph now has modules located in the correct directory. * Improve some graph drawing aspects (node margins, colours for module graphs, etc.). -- Ivan Lazar Miljenovic Ivan.Miljenovic(a)gmail.com IvanMiljenovic.wordpress.com

2 4

Fwd: [Haskell-cafe] Haskell for Physicists
by Alberto G. Corona 03 Oct '09

03 Oct '09

I´m a physicist, so I think they would be attracted by something like Haskell: high level physics modelling at Fortran speeds Haskell: mathematics beyond numerical calculus 2009/9/30 <edgar(a)ymonad.com> Hi, > > I will give a seminar to physicists at USP (Universidade de São Paulo, > Brazil) university and they asked me for a good title, something that can > attract physicists. Anyone has some suggestions? (Will be > a seminar about the use of Haskell to substitute C or Fortran > in a lot of tasks, and how it can be used in some problems instead of > Matlab, Mathematica, etc.) > > Thanks, > Edgar > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe >

6 8

Alex problem, help needed!!!!!!!!! ASAP :)
by María Laura 03 Oct '09

03 Oct '09

Dear list, I am trying to make a compiler and we are having a hard time getting Alex to work. We have succeded to work out Alex using older version, but with the the 2.2 version we keep getting this error and we havent been able to figure it out. So this is our tokens definition: { module Lexico where import Alex } %wrapper "posn" $digit = 0-9 -- digits $alpha = [a-zA-Z] -- alphabetic characters tokens :- $white+ ; --Los espacios en blanco los omito \/\/.* ; --Lo que venga después de dos barras omito \/\*.*\*\/ ; --Lo que está entre las barras de comentario omito $digit+ { \p s -> TokenEntero p (read s) } \' [$alpha $digit \_]* \' { \p s -> TokenString p (read s)} [$digit]+\.[$digit]+ { \p s -> TokenDouble p (read s) } $alpha [$alpha $digit \_]* { \p s -> TokenVar p (read s) } And when we call the alexScanTokens "hello" we get this error: [TokenVar (AlexPn 0 1 1) "*** Exception: Prelude.read: no parse So we are concerned about the "TokenVar p (read s)" . Would that be the way to read a string? We just get this error with TokenVar and TokenString, whenever we call the function like: alexScanTokens 19 , for instance we dont get any problems and the token 19 is recognized ok. Please, any ideas on what am i doing wrong???????? regards, L _________________________________________________________________ Keep your friends updated—even when you’re not signed in. http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/so…

3 2

Comments requested: succ Java
by Curt Sampson 02 Oct '09

02 Oct '09

No, it's not quite what it sounds like. :-) Stuart Halloway recently posted a series of blog entries entitled "Java.next"[1], discussing the benefits of four other languages that compile to JVM bytecode and interoperate with Java: Clojure, Groovy, JRuby, and Scala. I thought I'd put my oar in and write a parallel series comparing Haskell to these. I've finished a draft of the first posting, started on the third, and made a couple of notes on the second and fourth, and I thought I'd post the drafts[2] and solicit comments here. If you have time to read and comment, I'd greatly appreciate the help; feel free either to e-mail me privately or post here. Also feel free to forward this to anybody else you feel might be interested in commenting. I'll probably be posting these about one per week, starting some time next week. [1]: http://blog.thinkrelevance.com/2008/9/24/java-next-overview [2]: http://www.starling-software.com/en/blog/drafts/2009/09/27.succ-java-summar… cjs -- Curt Sampson <cjs(a)starling-software.com> +81 90 7737 2974 Functional programming in all senses of the word: http://www.starling-software.com

25 53

Haskell for Physicists
by edgar＠ymonad.com 02 Oct '09

02 Oct '09

Hi, I will give a seminar to physicists at USP (Universidade de São Paulo, Brazil) university and they asked me for a good title, something that can attract physicists. Anyone has some suggestions? (Will be a seminar about the use of Haskell to substitute C or Fortran in a lot of tasks, and how it can be used in some problems instead of Matlab, Mathematica, etc.) Thanks, Edgar

9 12

Introspection on types.
by Gregory Propf 02 Oct '09

02 Oct '09

Is there a way to tell, let's say, how many constructors there are for a type? Or do I need one of the haskell extensions I've read about?

3 3

ANN: json-b-0.0.4
by Jason Dusek 01 Oct '09

01 Oct '09

This version fixes defective handling of empty objects and arrays. http://hackage.haskell.org/package/json-b-0.0.4 Thanks to Dmitry Astapov for this fix. -- Jason Dusek

1 0

a boring parser
by Anatoly Yakovenko 01 Oct '09

01 Oct '09

i got annoyed with Parsec and wrote a much more boring parser which allows me to parse anything with any kind of matching i want. Its basically a combination of State and Error monads. So i can use a grep like parser that matches via a regular expression over a list of lines grep re = do vv::B.ByteString <- any let (_,_,_,rv) = (vv =~ re)::(B.ByteString,B.ByteString,B.ByteString,[B.ByteString]) case (rv) of [] -> throwError "no match" _ -> return $ rv > run (grep $ C.pack "(hello)") $ [C.pack "hello world"] Right (["hello"],[]) or use the same library to scan over a string by combining regular expressions regex re = do make $ \ st -> do case (B.null st) of True -> throwError "eos" _ -> do let (_,_,after,rv) = (st =~ re)::(B.ByteString,B.ByteString,B.ByteString,[B.ByteString]) case (rv) of [] -> throwError "no match" _ -> return $ (rv,after) > run (do aa <- regex $ C.pack "(hello)"; bb <- regex $ C.pack " (world)"; return (aa,bb) ) $ C.pack "hello world" Right ((["hello"],["world"]),"") or simply match integers in a list, or anything that is of type Eq > run (many1 $ eq 1) [1,1,1,2,3,4] Right ([1,1,1],[2,3,4]) i can define lt lt cc = do vv <- any case (vv < cc) of True -> return $ vv _ -> throwError "no match" and do > run (many1 $ lt 5 <|> eq 5) [1..10] Right ([1,2,3,4,5],[6,7,8,9,10]) here is the implementation module Parser( ParserM --type alias for the parser ParserM a b is over "stream" a and returns b , make --makes a parser from a matching function of type :: stream -> m (match_data,stream) --for example any is implemented via: --any :: ParserM [a] a --any = make $ \ ll -> -- case (ll) of -- (hh:tt) -> return $ (hh,tt) -- _ -> throwError "eos --matches and returns an element from a list, which makes any of type :: ParserM [a] a , any --matches any element from [a] type stream , eq --matches an equal element from [Eq] stream, trivialy implemented in terms of any --eq :: Eq a => a -> ParserM [a] a --eq cc = do -- vv <- any -- case (vv == cc) of -- True -> return $ vv -- _ -> throwError "no match , (<|>) --or operator, tries the left one then the right one , manyTill --collects the results of parser 1 until parser 2 succeeds , many1 --collects the results of the parser, must succeed at least once , many --collects the results of a parser , run --runs the parser ) where import Control.Monad.State.Lazy import Control.Monad.Error import Test.QuickCheck import Control.Monad.Identity import Prelude hiding (any) type ParserM a c = StateT a (ErrorT [Char] Identity) c make pp = do st <- get (rv,nst) <- pp $ st put $ nst return $ rv aa <|> bb = aa `catchError` \ _ -> bb manyTill :: ParserM a c -> ParserM a d -> ParserM a ([c],d) manyTill pp ee = do do dd <- ee return $ ([],dd) `catchError` \ _ -> do cc <- pp (ccs,dd) <- manyTill pp ee return $ (cc:ccs,dd) many1 pp = do rv <- pp rest <- many1 pp `catchError` \ _ -> return $ [] return $ rv : rest many pp = do many1 pp <|> return [] any :: ParserM [a] a any = make $ \ ll -> case (ll) of (hh:tt) -> return $ (hh,tt) _ -> throwError "eos" eq :: Eq a => a -> ParserM [a] a eq cc = do vv <- any case (vv == cc) of True -> return $ vv _ -> throwError "no match" lt cc = do vv <- any case (vv < cc) of True -> return $ vv _ -> throwError "no match" run pp dd = runIdentity $ runErrorT $ runStateT pp dd run' = flip run prop_MatchA = (Right ('a',"bc")) == (run' "abc" $ eq 'a') prop_MatchEOS = (Left "eos") == (run' "" $ eq 'a') prop_MatchNoMatch = (Left "no match") == (run' ("bcd") $ eq 'a') prop_MatchABC =(Right ('c',""))== (run' "abc" $ do eq 'a' eq 'b' eq 'c') prop_MatchA_C = (run' "abc" $ do eq 'a' eq 'd' <|> eq 'b' <|> any eq 'c') == (Right ('c',"")) prop_Or = (run' "abc" $ do { eq 'a' ; do { eq 'b' ; eq 'd' } <|> do { eq 'b' ; eq 'c' } }) == (Right ('c',"")) prop_UntilC = (Right (("",'c'),"")) == (run' ("c") $ manyTill any $ eq 'c') prop_Until1 ls = let rv = run' (ls ++ [1]) $ manyTill any $ eq 1 in case (rv) of Right ((ls,1),rest) -> (elem 1 ls) == False _ -> False prop_all1 ls = let rv = run' ([1,1,1] ++ ls) $ many1 $ eq 1 in case (rv) of Right (_,(1:_)) -> False Right ((1:1:1:_),_) -> True _ -> False

2 2