- Haskell-Cafe - Haskell.org

beginner question re example in Hutton's "Programming in Haskell"
by David Cabana 24 Feb '07

24 Feb '07

I have been trying to work through Graham Hutton's "Programming in Haskell", but have hit something of a snag in chapter 8.4. Hutton presents some sample code which I am trying to run, with no luck so far. Here is the code as I constructed it by gathering snippets presented across three pages. <code> module Main where -- as per Hutton page 75 type Parser a = String -> [(a, String)] -- page 75 return :: a -> Parser a return v = \inp -> [(v,inp)] -- page 76 failure :: Parser a failure = \inp -> [ ] item :: Parser Char item = \inp -> case inp of [ ] -> [ ] (x:xs)-> [(x,xs)] parse :: Parser a -> String -> [(a,String)] parse p inp = p inp -- page 77 p :: Parser (Char, Char) p = do x <- item item y <- item return (x,y) </code> When I tried to load this code, I got this error message: Ambiguous occurrence `return' It could refer to either `return', defined at /Users/joe/haskell/ parse2.hs:8:0 or `return', imported from Prelude OK. My reasoning was that Hutton took the trouble to define return, so I decided to use the local definition instead of the one in the Prelude. I changed the last line from "return (x,y)" to "Main.return (x,y)". The new error message is worse: Couldn't match expected type `Char' against inferred type `[(Char, String)]' In the expression: x In the first argument of `return', namely `(x, y)' In the expression: return (x, y)</blockquote> Hutton provided explicit type signatures, so I did not expect type issues. I decided to take another approach. The book has a website that lists errata and provides code listings. The code for chapter eight is at http://www.cs.nott.ac.uk/~gmh/Parsing.lhs When I read Hutton's code, I noticed that he begins by importing Monad. The code I list above is from chapter 8, pages 75-77 of the book. Monads have not yet been mentioned. The book's index shows that monads aren't mentioned till page 113, in chapter 10. I also notice that in his code, Hutton makes repeated use of a symbol P whose meaning I do not know. What do I have to do to make this code work? I know I can use Hutton's code from the website, but I expected the code presented in the book to work, or the code on the website to restrict itself to what has been discussed in the book. Am I missing something here? Thanks, David Cabana

2 1

Re: [Haskell-cafe] A "real" Haskell Cookbook
by P. R. Stanley 23 Feb '07

23 Feb '07

>>As a tip for anyone involved in writing and publishing scientific >>materials on the web, unless the maths is either written without any >>funny symbols or, better still, typeset in latex, it is not >>accessible to a screen-reader. > >I was under the impression that modern screen readers could pronounce >Unicode characters by looking up their name. I.e., your   would >get read as 'Non-breaking space' (perhaps a bad example, this one >wouldn't want to be read out due to its abuse as a layout tool, which >would make reading old pages very awkward). > >I don't see how images are going to be much better? I suppose <math> >images do, on MediaWiki, have an alt text which is their LaTeX, but >I'd hate to have to have that read to me. > I think latex is the perfect solution to the problem. It is > perhaps the only 100 percent accessible medium available right now. > It doesn't require any special software to read. All it needs is a > simple text editor. >Latex is a well-established tool/medium in the world-wide scientific >community and therefore its inclusion in the Haskell wikibook or any >other scientific document along with the unicode characters and >image files would be potentially beneficial to everyone. Paul >If you're interested in talking to the authors of the wikibook, >subscribe to the wikibook(a)haskell.org mailing list. > >-- >-David House, dmhouse(a)gmail.com

1 0

Re: [Haskell-cafe] A "real" Haskell Cookbook
by P. R. Stanley 23 Feb '07

23 Feb '07

Andrew and co, I have a big fat apology to make. It was actually the wikipedia page on Lambda Calculus where there are many occurrences of special symbols. The Wikibook, as far as I've explored, is fully accessible. Nonetheless, I'd still be more than happy to make my contributions in any way I can. I am a beginner to functional programming and so I could play a useful role in ensuring the text takes nothing for granted with regard to the fundamental principles and beyond. As a tip for anyone involved in writing and publishing scientific materials on the web, unless the maths is either written without any funny symbols or, better still, typeset in latex, it is not accessible to a screen-reader. One small point about the Haskell source code is that the column indentations do not always appear in the screen reader browse buffer. Of course, there are ways of getting around the issue and if the author of the Wikibook is interested we could discuss the details off list. Cheers, Paul At 16:22 22/02/2007, you wrote: >Can you guys point out an example of this in the wiki book? I've never >heard of this accessibility issue. I may even be willing to go back >through and modify the wikibook to make it more accessible of you can >explain what you're referring to. Accessibility is important to me. > >On 2/21/07, P. R. Stanley <prstanley(a)ntlworld.com> wrote: >>and can I please ask anyone thinking of using special symbols to >>resist the temptation. >>Symbols such as the &160 used liberally in the Haskell wikibook are >>totally invisible to screen readers. >>I would be happy to proof read any document before it goes to the >>wikibook to ensure it's fully accessible to screen readers. >> >>Regards, >>Paul >> >>At 03:17 22/02/2007, you wrote: >> >I made a preliminary page, and fleshed out some of the >> >headers/sub-headers on the wiki page for a good Haskell Cookbook >> >(aka NOT a PLEAC clone). Please contribute and/or fix the examples >> >and explanations so we can make a really nice Cookbook for newbies. :) >> > >> > http://haskell.org/haskellwiki/Cookbook >> > >> >_______________________________________________ >> >Haskell-Cafe mailing list >> >Haskell-Cafe(a)haskell.org >> >http://www.haskell.org/mailman/listinfo/haskell-cafe >> >>_______________________________________________ >>Haskell-Cafe mailing list >>Haskell-Cafe(a)haskell.org >>http://www.haskell.org/mailman/listinfo/haskell-cafe

4 3

process
by h. 23 Feb '07

23 Feb '07

Hello, I need to interact with some other program, and wrote the following code: module Main where import System.Process import System.IO main :: IO () main = do putStrLn "Running proc9..." (inp,out,err,pid) <- runInteractiveProcess "prog1" [] Nothing Nothing hSetBuffering inp LineBuffering hSetBuffering out LineBuffering hSetBuffering err LineBuffering hPutStrLn inp "1" a <- hGetLine out hPutStrLn inp a a <- hGetLine out waitForProcess pid putStrLn a But it does not work as I expected. As long as there is no need to put some input after having received some output it is no problem, but real interaction seems not possible. Is there a solution to work with the Handles as with stdin,... ? using ghc 6.6 on Windows Thanks in advance for your help.

8 13

exists . a psuedo-standard non-empty list module
by oleg＠pobox.com 23 Feb '07

23 Feb '07

Nicolas Frisby wrote: > Some of the code from the previous wiki link, type-level decimal > numbers I would rather advice against type-level decimal numbers, if we are looking at a lightweight solution. The [complete] code at http://www.haskell.org/haskellwiki/Non-empty_list is Haskell98! It also provides for a head and tail functions -- which are total and so raise no errors. Incidentally, inserting NList into the existing Safe.List does not seem like a good match as NList critically relies on being in a separate module with a limited export. NList constitutes a security kernel, whose invariants are to be described explicitly (and hopefully, automatically proven at some point). For that reason, it is beneficial to keep module NList separate and compact. It could of course be placed within Safe.* hierarchy, etc -- as a separate module. Neil Mitchell wrote: > Fortunately the problem of pattern match errors is being tackled by at > least two projects: > > * Catch: http://www-users.cs.york.ac.uk/~ndm/projects/catch.php > * ESC-Haskell: http://www.cl.cam.ac.uk/~nx200/research/escH-hw.ps I would submit the approach of the lightweight static capabilities (cf above Wiki link) to be counted as the third project in that area. The latter has an advantage that it is available in Haskell right now and requires no extra tools.

3 2

exists . a psuedo-standard non-empty list module
by Nicolas Frisby 23 Feb '07

23 Feb '07

I don't particularly like using fromJust or head, and there's been plenty of discussion on those issues. For the cases where it makes good sense to do so, I'm about to write a module for non-empty lists that looks to strike a balance between usability and static checks. But before I re-invent the wheel, I was wondering if there's already a popular module along those lines. I would also appreciate references to your favorite discussion about uses of head and other unsafe functions or various Oleg posts showing how they're all unnecessary. I'll find these on my own, but I would also like to know which ones strike a chord with the community. Thanks for your time. -Nick

4 5

Map list of functions over a single argument
by Paul Moore 23 Feb '07

23 Feb '07

I'm after a function, sort of equivalent to map, but rather than mapping a function over a list of arguments, I want to map a list of functions over the same argument. The signature would be [a -> b] -> a -> [b], but hoogle didn't come up with anything. It seems like an obvious analogue of map, so I'm pretty sure I'm missing something (otherwise, I'd just write it myself, and not bother :-)) Can anyone point me in the right direction? Thanks, Paul.

11 13

Re: [Haskell-cafe] TFP 2007: Registration and Program
by TFP 2007 23 Feb '07

23 Feb '07

>> >> Dear Colleagues, >> >> You may now resgister for TFP 2007! TFP 2007 will be held April 2-4, >> 2007 in New York City, USA. > April 2 is the first night of Passover. This is not one of those Your point is well taken. It is very unfortunate that the overlap you have pointed out exists. The organizing committee had to consider _many_ variables including, among others, the availability of a venue and of hotels for participants in NYC. Sometimes the logistics of a situation force your hand. ************************************************************************************ Dr. Marco T. Morazan TFP 2007 Program Committee Chair http://cs.shu.edu/tfp2007/

1 0

Prime finding
by Ruben Zilibowitz 23 Feb '07

23 Feb '07

I see that there has been some discussion on the list about prime finding algorithms recently. I just wanted to contribute my own humble algorithm: primes :: [Integer] primes = primesFilter 1 [2..] primesFilter :: Integer -> [Integer] -> [Integer] primesFilter primorial (n:ns) | (gcd primorial n == 1) = n : (primesFilter (primorial*n) ns) | otherwise = primesFilter primorial ns Comparing it to some of the algorithms in: http://www.haskell.org/pipermail/haskell-cafe/2007-February/022765.html It seems to perform reasonably well. It also has the advantage of being quite short. Ruben

3 5

HList, hOccurs and MonadReader
by Marc Weber 23 Feb '07

23 Feb '07

Having the module given below I can't see why using printAndRerun l1 printAndRerun2 l2 but not printAndRerun l1 printAndRerun l2 ? They only differ in their name. Can you point me in the right direction? ------------------------------------------------ {-# OPTIONS -fglasgow-exts #-} {-# OPTIONS -fallow-undecidable-instances #-} {-# OPTIONS -fallow-incoherent-instances #-} {-# OPTIONS -fallow-overlapping-instances #-} module HListTest.MonadReader where import HList hiding (liftIO) import Monad import Control.Monad.Reader printAndRerun = do v <- liftM hOccursFst ask liftIO $ print (v :: Int ) v2 <- liftM hOccursFst ask return (v2 :: String) printAndRerun2 = do v <- liftM hOccursFst ask liftIO $ print (v :: Int ) v2 <- liftM hOccursFst ask return (v2 :: String) a = (2 :: Int) b = "str" l a b = HCons a (HCons b HNil) --- l1 :: ( HCons Int (HCons String HNil)) l1 = l a b l2 = l b a printBoth f l = runReaderT f l >>= print printBoth2 l = do print (hOccurs l :: Int) print (hOccurs l :: String) hlistMonadReaderTest = do print "hlistMonadReaderTest" printBoth printAndRerun l1 -- printBoth printAndRerun l2 -- < this doesn't work but the next line ? Where is the difference between printAndRerun printAndRerun2 ? printBoth printAndRerun2 l2 -- ^ printBoth2 l1 -- here is no trouble.. printBoth2 l2 main = hlistMonadReaderTest ------------------------------------------------ when not commenting the line above I'm getting this error: HListTest/MonadReader.hs|35| 26: Couldn't match expected type `Int' against inferred type `[Char]' Expected type: HCons Int (HCons [Char] HNil) Inferred type: HCons [Char] (HCons Int HNil) In the second argument of `printBoth', namely `l2' In the expression: printBoth printAndRerun l2 Thanks Marc

2 2