- Beginners - Haskell.org

Exercise of "Programming with Arrows"
by Thiago Negri 16 Oct '13

16 Oct '13

The paper "Programming with Arrows" of John Hughes gives some exercises to do [1]. I'm trying to solve it and would like to receive a feedback if I'm doing it right or not before reading the rest of the paper. I didn't find the answers in the internet (if someone could point me to it, please do so). Exercise 2 (section 2.5) is asking to create a Stream Processor that can map more than one output per input (e.g. 3 outcomes for a single consume of the stream). The paper says that implementing "first" will be tricky, and it really is. I've came up to the solution listed below, *is it right?* * * * * module SP where import Prelude hiding (id, (.)) import Control.Category import Control.Arrow data SP a b = Put b (SP a b) | Get (a -> SP a b) runSP :: SP a b -> [a] -> [b] runSP (Put b s) as = b:runSP s as runSP (Get k) (a:as) = runSP (k a) as runSP (Get k) [] = [] compose :: SP b c -> SP a b -> SP a c compose (Put a s) g = Put a (compose s g) compose (Get k) (Put a s) = compose (k a) s compose f (Get k) = Get (\a -> compose f (k a)) instance Category SP where id = arr id (.) = compose instance Arrow SP where arr f = Get (\a -> Put (f a) (arr f)) first (Put a s) = Get (\(a', c) -> Put (a, c) (delayed (a', c) s)) first (Get k) = Get (\(a, c) -> firstWithValue (k a) c) delayed :: (a, c) -> SP a b -> SP (a, c) (b, c) delayed (a, c) (Put b s) = Put (b, c) (delayed (a, c) s) delayed (a, c) (Get k) = firstWithValue (k a) c firstWithValue :: SP a b -> c -> SP (a, c) (b, c) firstWithValue (Put a s) c = Put (a, c) (firstWithValue s c) firstWithValue (Get k) _ = Get (\(a, c) -> firstWithValue (k a) c) input :: [(String, String)] input = [("a1", "a2"), ("b1", "b2"), ("c1", "c2"), ("d1", "d2")] myArrow :: SP (String, String) (String, String) myArrow = (delay "db1" >>> delay "da1") *** (delay "db2" >>> delay "da2") delay :: a -> SP a a delay b = Put b (arr id) main :: IO () main = let output = runSP myArrow input in mapM_ f output where f (a, b) = putStrLn $ "(" ++ show a ++ ", " ++ show b ++ ")" * * The output of "main" is: *SP> main ("da1", "da2") ("da1", "db2") ("da1", "a2") ("db1", "a2") ("a1", "a2") ("b1", "b2") ("c1", "c2") ("d1", "d2") [1] http://www.cse.chalmers.se/~rjmh/afp-arrows.pdf

2 13

Problems with cabal install hakyll
by Britt Anderson 16 Oct '13

16 Oct '13

I have a fresh installation of ghc 7.6.3-1 and cabal 1.16. From this I was able to cabal update and then cabal install diagrams and pandoc, but when trying to install hakyll (either version 4 or 3) I get a lot of dependency conflicts. Can anyone give me some suggestions on how to proceed in this case? Thank you. Below follow the messages from cabal install --dry-run *** Trying version 4 *** cabal install --dry-run hakyll Resolving dependencies... cabal: Could not resolve dependencies: trying: hakyll-4.4.1.0 trying: hakyll-4.4.1.0:+checkexternal trying: http-conduit-1.9.5.1/installed-1fe... rejecting: tls-extra-0.6.6/installed-302... (conflict: tls-extra => cryptohash==0.11.1/installed-ae4..., hakyll => cryptohash>=0.7 && <0.11) rejecting: tls-extra-0.6.6, 0.6.5, 0.6.4, 0.6.3, 0.6.1, 0.6.0, 0.5.1, 0.5.0, 0.4.7.1, 0.4.7, 0.4.6.1, 0.4.6, 0.4.5, 0.4.4, 0.4.3, 0.4.2.1, 0.4.2, 0.4.1, 0.4.0, 0.3.1, 0.3.0, 0.2.3, 0.2.2, 0.2.1, 0.2.0, 0.1.9, 0.1.8, 0.1.7, 0.1.6, 0.1.5, 0.1.4, 0.1.3, 0.1.2, 0.1.1, 0.1.0 (conflict: http-conduit => tls-extra==0.6.6/installed-302...) ***Trying version 3 *** cabal install --dry-run hakyll-3.5.3.0 Resolving dependencies... In order, the following would be installed: HUnit-1.2.5.2 (new package) List-0.5.1 (new package) blaze-html-0.5.1.3 (new version) bytestring-mmap-0.2.2 (new package) entropy-0.2.2.4 (new package) crypto-api-0.12.2.2 (new package) cryptohash-0.8.4 (new version) enumerator-0.4.20 (new package) attoparsec-enumerator-0.3.1 (new package) blaze-builder-enumerator-0.2.0.5 (new package) hashable-1.1.2.5 (new version) case-insensitive-1.1 (reinstall) changes: hashable-1.2.1.0 -> 1.1.2.5 hexpat-0.20.3 (new package) highlighting-kate-0.5.5 (reinstall) changes: blaze-html-0.6.1.1 -> 0.5.1.3, filepath-1.3.0.1 added lrucache-1.1.1.3 (new package) regex-tdfa-1.1.8 (new package) syb-0.3.7 (new version) hs-bibutils-5.0 (new package) json-0.7 (new package) pandoc-types-1.9.1 (new version) citeproc-hs-0.3.9 (new package) system-filepath-0.4.8 (new package) system-fileio-0.3.11 (new package) shakespeare-1.2.0.1 (new package) hamlet-1.1.7.3 (new package) tagsoup-0.12.8 (new version) texmath-0.6.4 (reinstall) changes: pandoc-types-1.12.2.3 -> 1.9.1, syb-0.4.1 -> 0.3.7 pandoc-1.9.4.5 +blaze_html_0_5 (new version) unix-compat-0.4.1.1 (new package) unordered-containers-0.2.3.3 (reinstall) changes: hashable-1.2.1.0 -> 1.1.2.5 zlib-enum-0.2.3 (new package) snap-core-0.9.4.1 (new package) snap-server-0.9.3.4 (new package) hakyll-3.5.3.0 (new package) Warning: The following packages are likely to be broken by the reinstalls: http-types-0.8.1 pandoc-1.12.0.2 http-conduit-1.9.5.1 yaml-0.8.5.1 semigroups-0.11 void-0.6.1 zlib-conduit-1.0.0 lens-3.9.2 force-layout-0.2 diagrams-contrib-0.7 diagrams-0.7.1.1 conduit-1.0.8 blaze-builder-conduit-1.0.0 MemoTrie-0.6.1 vector-space-0.8.6 vector-space-points-0.1.2.1 diagrams-core-0.7.0.1 diagrams-svg-0.8.0.1 diagrams-lib-0.7.1.1 active-0.1.0.7 semigroupoids-3.1 semigroupoid-extras-3.0.1 profunctor-extras-3.3.3.1 monoid-extras-0.3.2.2 dual-tree-0.1.0.4 groupoids-3.0.1.1 comonad-transformers-3.1 comonads-fd-3.0.3 bifunctors-3.2.0.1 comonad-3.1 profunctors-3.3.0.1 aeson-0.6.2.1 pandoc-types-1.12.2.3 Use --force-reinstalls if you want to install anyway.

3 5

Re: [Haskell-beginners] [Haskell-cafe] Haskell Group Dresden?
by Daniel Trstenjak 15 Oct '13

15 Oct '13

Hi Stephan, > I'm very much interested. Do you need any help organizing a first meeting? > > For what it's worth, we once had an event about Yesod at the local CCC > group[1]. well, normally I'm not the kind of guy who enjoys performing in front of a lot of people, so I'm trying to find a way to get in touch with some folks without getting a heart attack =). So looking at your yesod event, the perception to get filmed, isn't really helping here :D. I was thinking about something "more" informal or if someone else wants to talk, I have no issues to sit somewhere in the back ;). Ok, but perhaps we can work out something fitting. Greetings, Daniel

1 0

Haskell Group Dresden?
by Daniel Trstenjak 15 Oct '13

15 Oct '13

Hi, is there any interest in having something like a Haskell Group in Dresden? Perhaps I could say something about the lens library, which would be more of a practical talk, how to define lenses and working with lenses, traversals, isos and prisms. Greetings, Daniel

1 0

Re: [Haskell-beginners] monads do not fit together?
by Brandon Allbery 15 Oct '13

15 Oct '13

On Tue, Oct 15, 2013 at 5:00 AM, Kees Bleijenberg < k.bleijenberg(a)lijbrandt.nl> wrote: > I wonder, do you always have to create a monadtransformer if your program > uses 2 kinds of IO? > In one sense the answer is "yes"; but the real answer is "that's not how it works". There is only one IO. Monad transformers allow you to add functionality on top of it, but at some point you need to do actual I/O in the IO monad. One reason it's done this way is that CGIT could in theory be written such that you could swap out the raw IO monad for something that lets you do testing. -- brandon s allbery kf8nh sine nomine associates allbery.b(a)gmail.com ballbery(a)sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net

1 0

Re: [Haskell-beginners] monads do not fit together?
by Brandon Allbery 15 Oct '13

15 Oct '13

On Mon, Oct 14, 2013 at 10:58 AM, Kees Bleijenberg < k.bleijenberg(a)lijbrandt.nl> wrote: > I get the same message (Couldn't match expected type `CGIT IO a0' with > actual type `IO [String]'). I think I do understand where this message is > coming from. But what can I do to fix it? > For now, you just correct the type of runRefreshFirewall; it should be runRefreshFirewall :: String -> CGIT IO [String] If you are intending to run actual IO actions in the real thing, you'll need to use liftIO to "reach" the IO embedded in the CGIT IO a. -- brandon s allbery kf8nh sine nomine associates allbery.b(a)gmail.com ballbery(a)sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net

2 1

Re: [Haskell-beginners] Exercise of "Programming with Arrows"
by Kim-Ee Yeoh 14 Oct '13

14 Oct '13

On Tue, Oct 8, 2013 at 4:08 AM, Thiago Negri <evohunz(a)gmail.com> wrote: > I agree with "no-op assembly line", but when I'm using `first` on a processor, I want to process the first stream *only*. The second stream should remain as it was not touched, so future processors will receive the same sequence from the second stream. We are in violent agreement! > I mean, I think I need to guarantee that this definition holds: > > `g *** f` is the same as `first g >>> swap >>> first f >>> swap` Excellent! Let's turn to laws and definitions to be _precise_ in what we're saying. > If my implementation of `first` uses a real no-op assembly line for `c` (i.e., `arr id`), then I would lose the stream. To that end, consider this easy problem: using (1) the arrow laws that you know off the top of your head, and (2) the above definition of (***) can you show that g *** (arr id) = first g? Because arr id, as you noted, is a no-op. Do you see what's going on? -- Kim-Ee

1 0

Re: [Haskell-beginners] monads do not fit together?
by Kim-Ee Yeoh 14 Oct '13

14 Oct '13

On Mon, Oct 14, 2013 at 6:51 PM, Kees Bleijenberg < k.bleijenberg(a)lijbrandt.nl> wrote: > The error at the problem line is: parse error on input `<-' **** > > If I move the line above the ’if’ I get: Couldn't match expected type > `CGIT IO t0' with actual type `IO [String]'. **** > Looks like you're missing a "do". You're writing code like this: if blah then x <- foo bar x else quux when you need to write if blah then do x <- foo bar x else quux Recommended exercises: * can you see in your mind's eye the desugaring of do-syntax? * how would you write it in a single line, i.e. without using indentation a.k.a. "layout" -- hint: look up curly braces -- Kim-Ee

2 1

monads do not fit together?
by Kees Bleijenberg 14 Oct '13

14 Oct '13

The program belows reads the get fields 'un' and 'ip'. If these get fields both exist, the program modifies a firewall script. The problem is in cgiMain. The error at the problem line is: parse error on input `<-' If I move the line above the 'if' I get: Couldn't match expected type `CGIT IO t0' with actual type `IO [String]'. How can I extract values from the IO [String] monad in cgiMain? I don't want to alter the type of runRefreshFirewall. Kees module Main( main ) where import Network.CGI import Text.XHtml import Text.XHtml.Transitional import Data.Maybe runRefreshFirewall :: String -> String -> IO [String] runRefreshFirewall un ip = do return ["test" ] inputCgiOkPage :: String -> String -> [String] -> Html inputCgiOkPage un ip msgs = body << h1 << ("IP: " ++ (show ip)) undefinedFieldsPage :: Html undefinedFieldsPage = body << h1 << "Undefined Fields" cgiMain :: CGI CGIResult cgiMain = do maybeUn <- getInput "un" maybeIp <- getInput "ip" if ((isJust maybeUn) && (isJust maybeIp)) then msgs <-runRefreshFirewall (fromJust maybeUn) (fromJust maybeIp) -- problem output $ renderHtml $ inputCgiOkPage (fromJust maybeUn) (fromJust maybeIp) msgs else output $ renderHtml undefinedFieldsPage main :: IO () main = runCGI $ handleErrors cgiMain

1 0

Control.Exception.bracket-like for MonadIO ?
by Olivier Iffrig 14 Oct '13

14 Oct '13

Hello, I'm trying to build an application with a monad encapsulating its state, environment, ... For the sake of simplicity, let's say data MyState type Main = StateT MyState IO But I would like to use something like bracket with my Main monad, and I did not find how to to it. I'm looking for something like bracket' :: MonadIO m => m a -> (a -> m b) -> (a -> m c) -> m c or in my case myBracket :: Main a -> (a -> Main b) -> (a -> Main c) -> Main c Searching for that on hoogle did not give me anything interesting... Did someone encounter this problem, and have a solution ? -- Olivier

2 2