- Beginners - Haskell.org

Making sense of currying in the context of the Hask Category
by Lino Rosa 29 Sep '12

29 Sep '12

Hi, I'm a Haskell newbie and I couldn't quite make sense of how currying maps to the the Hask Category. How would I map, for instance (+) to a Hask 'arrow'? If objects are types on Hask, then would a -> a -> a be the first object on this chain? In that case, for the first arrow, would I have as many arrows as there are possible partial applications on this type? In other words, would I have (+) 1, (+) 2, (+) 3 ... all those transitioning to the second object ( a -> a ) Or, do I have ONE arrow only, like (+) a ? In either case, what happens after 'm left with the object a -> a? What function (arrow) mutates it to the final value 'a'? That's the function resulting from the previous partial application of (+), but that fuction only exists at run time, after you apply the first one. I guess the question is, if you'd have to write a diagram for this, what would you write beside each object and beside each arrow?

5 7

confused by <-
by Bryce Verdier 29 Sep '12

29 Sep '12

I'm trying to play with a Yesod application. I had working code before, but tried stretching myself a bit and adding in some error checking. This code worked: getGoogR = do body <- simpleHttp "http://www.google.com" jsonToRepJson $ object ["response" .= (showDigest $ sha1 body)] This is the new code that is trying to use try (imported from Control.Exception) getGoogR = do body <- try (simpleHttp "http://www.google.com") :: IO (Either SomeException Data.ByteString.Lazy.Internal.ByteString) case body of Left _ -> jsonToRepJson $ object ["response" .= ( show "ERROR: NO DATA FOR ONE REASON OR ANOTHER")] Right val -> jsonToRepJson $ object ["response" .= (showDigest $ sha1 val)] and I get this error: Couldn't match expected type `IO b0' with actual type `GHandler sub0 master0 RepJson' In the expression: jsonToRepJson $ object ["response" .= (show "ERROR: NO DATA FOR ONE REASON OR ANOTHER")] In a case alternative: Left _ -> jsonToRepJson $ object ["response" .= (show "ERROR: NO DATA FOR ONE REASON OR ANOTHER")] In a stmt of a 'do' block: case body of { Left _ -> jsonToRepJson $ object ["response" .= (show "ERROR: NO DATA FOR ONE REASON OR ANOTHER")] Right val -> jsonToRepJson $ object ["response" .= (showDigest $ sha1 val)] I'm sure that my problem has a simple fix, but I just don't know what it is. I will be happy to share any other code if people ask, I'm just giving the short version for brevities sake. Thanks in advance, Bryce

5 13

FRP
by Christopher Howard 28 Sep '12

28 Sep '12

Hi. I'm trying to understand what Functional Reactive Programming is, or, more properly, what distinguishes it from "regular" functional programming, and what it would look like if I were to program in a "FRP Style", i.e., without some mysterious FRP module hiding the details. After reading a bunch of links and even looking at some source code I'm still not clear on this. If I program in an event-based style, and have data structures updating over time, with some data structures dependent on others for their values, and I use a functional language, is that FRP? Or is there some essential element I'm missing here? -- frigidcode.com indicium.us

5 9

Lazy variant of sequence (or other way to approach problem)
by Nathan Hüsken 27 Sep '12

27 Sep '12

Hey, In my (SDL based) haskell program, I do: events <- liftM ( takeWhile (/= NoEvent)) $ sequence $ repeat pollEvent The execution of this never returns, I am guessing that is because sequence evaluation never stops. But if sequence would be lazy (and assuming pollEvent returns NoEvent at some point) this should stop, should it not? Is there a lazy variant of sequence? Or am I missing something here completely? Thanks! Nathan

3 4

Picking a random element from a Map
by Harald Bögeholz 27 Sep '12

27 Sep '12

Hi, how do I (quickly) pick a random element from a `Data.Map.Map`? Right now I am using `head $ Map.keys p` to get the first element, but I'd like to replace this by a random element. I know I have to drag around the state of the random generator somewhere, but if I have one I am wondering what's the best way to use one to take a pick from the map. Thanks -- Harald Bögeholz <bo(a)ct.de> (PGP key available from servers) Redaktion c't Tel.: +49 511 5352-300 Fax: +49 511 5352-417 http://www.ct.de/ int f[9814],b,c=9814,g,i;long a=1e4,d,e,h; main(){for(;b=c,c-=14;i=printf("%04d",e+d/a),e=d%a) while(g=--b*2)d=h*b+a*(i?f[b]:a/5),h=d/--g,f[b]=d%g;} (Arndt/Haenel) Affe Apfel Vergaser /* Heise Zeitschriften Verlag GmbH & Co. KG * Karl-Wiechert-Allee 10 * 30625 Hannover * Registergericht: Amtsgericht Hannover HRA 26709 * Persönlich haftende Gesellschafterin: Heise Zeitschriften Verlag * Geschäftsführung GmbH * Registergericht: Amtsgericht Hannover, HRB 60405 * Geschäftsführer: Ansgar Heise, Dr. Alfons Schräder */

2 2

Mapping list over datatype using Traversable and State monad.
by Dmitriy Matrosov 26 Sep '12

26 Sep '12

Hi. I have a data type data Line a = Line [a] [a] which groups elements into "ordered" (first list) and "other" (second list) ones. And i want a functions representing Line, which actually has two heads (one from 1st list and one from 2nd), as single-headed. In other words, i want to map a list over Line (preserving Line structure), i.e. implement a function with type f :: [a -> b] -> Line a I have done this using Traversable import Data.Monoid import qualified Data.Foldable as F import qualified Data.Traversable as T import Control.Applicative import Control.Monad.State instance Functor Line where fmap f (Line xs ys) = Line (map f xs) (map f ys) instance F.Foldable Line where foldMap f (Line xs ys) = (F.foldMap f xs) `mappend` (F.foldMap f ys) instance T.Traversable Line where traverse f (Line xs ys) = Line <$> (T.traverse f xs) <*> (T.traverse f ys) and my function (which actually used for adding separators (sp ++) into (Line String), and i don't want to add separator before the first element, regardless of whether first element is "ordered" or "other") looks like inlineSeps :: (a -> a) -> Line a -> Line a inlineSeps g = fst . flip runState (id : repeat g) . T.mapM f where f x = do (f : fs) <- get put fs return (f x) It works, but i'm not sure whether using state monad here is good? And whether this is good solution for such problem at all?

2 3

alut initialization
by Christopher Howard 24 Sep '12

24 Sep '12

Hi. I'm trying to learn ALUT/OpenAL modules so I can have sound in my app. I'm a bit confused by the alut module interface, because it uses this alutRun function that is different than what is in the actual alut library documentation. I thought I figured it out, but apparently not because this simple program doesn't work: code: -------- import Sound.ALUT main = do putStrLn $ "alutAPIVersion: " ++ alutAPIVersion runALUT "" [] (\x y -> return ()) mimes <- get bufferMIMETypes mapM putStrLn mimes *Main> main alutAPIVersion: 1.1 *** Exception: user error (bufferMIMETypes: The operation was invalid in the current ALUT state) -------- Is there anyone familiar with the module that can point me in the right direction? -- frigidcode.com indicium.us

1 0

elegant handling of maybe and non-maybe values
by Christopher Howard 21 Sep '12

21 Sep '12

I have a code situation similar to this simplified one: Let's say that I've got a value v that is a Maybe type, and values x, y, and z that are not. x, y, and z are dependent on the value inside v. So I must provide default values for x, y, and z. An additional complication is that every calculation returns a new (possibly different) value for v. So, I could do something like this: code: -------- let (v', x) = case v of Nothing -> (Nothing, defaultValueOfX) Just vValue -> f vValue in let (v'', y) = case v' of Nothing -> (Nothing, defaultValueOfY) Just vValue -> g vValue in let (v''', z) = case v'' of Nothing -> (Nothing, defaultValueOfZ) Just vValue -> h vValue in -------- (f, g, h represent the calculations I am performing.) In the end, I need the values of x, y, and z, and the last v value. Obviously, the above solution is not very elegant, especially if I add a few more variables. My other thought was some kind of fold operation, where I store the functions (calculations) and the default values in a list, and then fold over the list, applying subsequent values of v or returning default values, as appropriate. But this seems rather clunky, especially if I want to attach the values in the new list to variable names afterwards. Is there a better approach? -- frigidcode.com indicium.us

3 2

an observation about Haskell vs. Python
by Dennis Raddle 21 Sep '12

21 Sep '12

As a somewhat-newbie haskell user and longtime Python user, what I have observed is this. Haskell creates compile-time error messages that are somewhat hard to understand for beginners. Python (or any scripting language) creates run-time bugs that are hard to understand. One reason for the weird (to a beginner) compile errors in Haskell is its expressivity -- almost any sequence of identifiers could potentially mean something, and if you make a mistake, the compiler is sure to find some "weird" way to interpret it. But Python suffers from a similar problem -- it's not as expressive a language, but it is very permissive, not insisting on type correctness, order of arguments, or any of a number of things so that the programmer can write a program that compiles with no errors -- but has strange run-time bugs. I'll take Haskell. I'm a bit OCD about getting the bugs out of my programs, and Python just opens up too many holes for me to relax with it. Dennis

6 7

an observation about Haskell vs. Python
by Dennis Raddle 21 Sep '12

21 Sep '12

Thanks. I'm okay with the status quo.. I just see it as a tradeoff. You are giving up something to get something. In Python you give up any kind of compile-time type checking, you give up the safety of immutable data, and you get a whole lot of automatic type conversions and brief ways to express certain algorithms. In Haskell you give up easy-to-comprehend error messages and mutable data, and get back in reward a lot of reassurance that your program does what you meant and expressivity. (I realize Haskell has mutable data but it's not like Python's) On Wed, Sep 19, 2012 at 7:43 PM, Ertugrul Söylemez <es(a)ertes.de> wrote: > > Notice: You probably forgot to apply `sin' to an argument. > > However, I think that no work is done on that, and there is another > possible path: An average Haskell tutorial should always include a > section on understanding error messages. In fact the latest issue of > The Monad Reader [1] has an article on that by Jan Stolarek. > > [1]: http://themonadreader.files.wordpress.com/2012/08/issue20.pdf > > > Greets, > Ertugrul > > -- > Not to be or to be and (not to be or to be and (not to be or to be and > (not to be or to be and ... that is the list monad. > > _______________________________________________ > Beginners mailing list > Beginners(a)haskell.org > http://www.haskell.org/mailman/listinfo/beginners > >

3 2