- Haskell-Cafe - Haskell.org

Really Simple explanation of Continuations Needed
by Mark Spezzano 05 Oct '11

05 Oct '11

Hi, Can someone please give me a _lucid_ and _simple_ explanation of exactly how continuations can be used in Haskell? I've already had a look at most of the tutorials and explanations on the web, but I'm still confused. Continuations and CPS have me baffled. (I have most of the Haskell textbooks and even these are sketchy on Continuations) I don't understand the notion of the Cont monad and how it can be used for multitasking, backtracking and interrupting computations. I understand that functions take in a (continuation) function that represents the work remaining to do, but al of the explanations on the web and in technical papers seems to trip over themselves in explaining the fundamentals to a CPS-newbie. If anyone could explain such concepts to me in unambiguous, clear English then this would be very helpful. Thanks in advance for your help, Mark

6 7

categories in Haskell and module names
by Roman Beslik 05 Oct '11

05 Oct '11

Hello. IMHO there are 2 ways to define categories in Haskell: W0) the category of types and functions ("base.Data.Functor" and company belongs to it); W1) the class "base.Control.Category.Category". (Defining a category where the class of objects is a type seems impossible in Haskell.) But these ways are mingled in "base". E.g. "Control.Applicative" (W0) and "Control.Category" (W1) are neighbors. Oddly, "Data.Functor" (W0) is under "Data". I believe that the whole library "category-extras" belongs to W1. I suggest to split these ways under distinct names, e.g.: W0) Category.Function (@hom == (->)@); W1) Category.Hom ("hom" is an abstract type constructor). Maybe both should be under "Mathematics".

1 0

Munich Haskell Meeting
by Heinrich Hördegen 05 Oct '11

05 Oct '11

Dear all, last Thursday's get-together for Haskell users in Munich was a success. About 12 people meet at Cafe Puck where we spent a nice evening. The overall opinion was, that gatherings of this kind should be held on a more regular basis. Therefore, I'd like to announce the following page: http://www.haskell-munich.de If you can make it to Munich, please feel invited to join us at one of the given dates. We would highly appreciate your company. If you have suggestions about the page or anything else, please contact me. Have a nice day, Heinrich -- -- kontakt(a)funktional.info www.funktional.info --

1 0

Is it possible to represent such polymorphism?
by Du Xi 05 Oct '11

05 Oct '11

--I tried to write such polymorphic function: expand (x,y,z) = (x,y,z) expand (x,y) = (x,y,1) --And it didn't compile. Then I added a type signature: expand::a->b expand (x,y,z) = (x,y,z) expand (x,y) = (x,y,1) --It still didn't compile. I think the reason is that the following is disallowed: f::a->b f x = x --Is it possible to get around this and write the "expand" function? Of course, x and y may be of different types

16 25

Haddocok in Haskell Platform 2011.2.0.1?
by . 05 Oct '11

05 Oct '11

Hi, I just installed a new computer, downloaded ghc 7.0.3 and HP 2011.2.0.1 from http://hackage.haskell.org/platform/linux.html as proposed there, and I can not install haddock. It says it requires ghc >=7.2 and <=7.4. Is that intended? Cheers, Christian

2 2

DSL for data definition (e.g. compiling Haskell type defs into Google's protocol buffers type defs)
by Karel Gardas 05 Oct '11

05 Oct '11

Hello, I'm trying to find out if it's possible to use Haskell data type definition capability to define types and compile defined types into other languages, for example into Google's protocol buffers data definition language. So basically speaking I'm thinking about using Haskell sub-set as a data-definition DSL together with some functions which will generate some code based on supplied defined data types. My idea is: data Person = Person { id :: Int , name :: String , email :: Maybe String } deriving (Show, Data, Typeable) emit_proto Person 1 where emit_proto is function which will translate Person data type definition into Google's proto language (the 1 is index from which start to index type's fields) by traversing data type definition and translating all its children plus do some header/footer generation etc: message Person { required int32 id = 1; required string name = 2; optional string email = 3; } I've looked for something like that and found SYB papers which works on top of data instance (i.e. actual data, not data type). I also found JSON lib which again works on top of data and not data type. I've tried to look into Data.Typetable etc, but have not found function which will print data type's field name and field type name (although JSON lib seems to use field name for JSON generation so I'll need to investigate this more). I've tested `typeOf' function and it's quite useful, but its limitation is that it's not working on ADT name: data Color = RED|GREEN|BLUE *Main> typeOf Color <interactive>:1:8: Not in scope: data constructor `Color' *Main> typeOf RED Main.Color and I would need that in order to translate Color defined above into enum like: enum Color { RED = 0; GREEN = 1; BLUE = 2; } My question is: do you think I'm looking into good direction (i.e. Data/Typeable) or do you think I'll need to use something different for data definition DSL (Template Haskell?, or impossible in Haskell so write my own language with full parser? etc?) Thanks for any idea or opinion on this! Karel

8 9

ANN: hledger 0.16
by Simon Michael 05 Oct '11

05 Oct '11

I'm pleased to announce version 0.16 of the hledger packages. This is a stability/bugfix/polish release (which may become the pattern for even-numbered releases in future.) hledger is a library and set of user tools for working with financial data (or anything that can be tracked in a double-entry accounting ledger.) It is a haskell port and friendly fork of John Wiegley's Ledger. hledger provides command-line, curses and web interfaces, and aims to be a reliable, practical tool for daily use. Given a plain text file describing transactions of money or any other commodity, it will print the chart of accounts, account balances, or just the transactions you're interested in. It can also help you record new transactions, or convert CSV data from your bank. Home: http://hledger.org To install: cabal update, cabal install hledger. If you have trouble, let me know or watch the mail list/bug tracker for updates. Optional add-ons: hledger-web hledger-vty hledger-chart hledger-interest. Pre-built platform-specific binaries, which are time-consuming to make and support, are provided on request to anyone who donates. This is a great way to give back and support the project! http://hledger.org/DOWNLOAD.html Release notes for 0.16: * cli: strip the -- when calling add-on commands, so their options work (#64) * cli: hledger ADDON --version now shows add-on command's version * cli: only the add and web commands auto-create the journal file * cli: give a non-confusing error if LEDGER_FILE contains a literal tilde * add: clearer prompts, more validation, use . to end also * add: use unix line endings consistently, avoiding parse error on windows (#51) * add: avoid excess whitespace between transactions (#46) * balance: ledger compatibility fix: don't elide parent accounts with multiple displayed subaccounts * convert: always order converted transactions by date * convert: rename currency -> base-currency, in-field, out-field -> amount-in-field, amount-out-field * convert: give an error, not a zero when date or amount-in-field/amount-out-field parsing fails * register: show more useful range of intervals with --empty and a query pattern * print, web: always show both dates, ignoring --effective (#42) * web: production builds (the default with cabal) have all web content embedded (dev builds use ./static/) (#63) * web: update to yesod 0.9 * web: obey at least some of the general reporting options, like --cost * web: adjust the default base url when a custom port is specified * web: prevent an infinite redirect when custom base url has a trailing slash * web: fix "not:'multi word'" patterns * web: hide old title and search form when adding/editing * web: adjust --help to indicate command-line arguments are not expected * web: don't bother running cli unit tests at startup Stats: - Release contributors: Simon Michael - 30 days, 100 commits, 4 end-user features and 18 end-user bugfixes since last major release - 219 unit & functional tests and 55% unit test coverage (hledger, hledger-lib packages) - 7632 lines of code (all packages) Best, -Simon

1 1

Fwd: LZMA for Haskell?
by Paulo Pocinho 05 Oct '11

05 Oct '11

(Thought it would be better to put this in the haskell-cafe list). Hello list. I'd like to use LZMA to create/extract archives and compare CRC from archived files. What do I need to use LZMA in Haskell? There is a page called "Library/Compression" in the wiki [1] but it is from 2005, using LZMA version 4.06. Besides being old, I can only find the LZMA SDK version 9.20 for download. The only related package I can find on hackage is lzma-enumerator [2]. Trying to install with cabal requires an old header named lzma.h that is no longer available. [1] http://www.haskell.org/haskellwiki/Library/Compression [2] http://hackage.haskell.org/package/lzma-enumerator

2 1

Composing a list of Enumeratees into an Enumerator using ($=)
by Román González 05 Oct '11

05 Oct '11

Hey guys, Right now I'm facing with a type problem that is really nasty, I want to compose a list of enumeratees using the ($=) operator to create a new enumerator. Whenever I'm trying to use the foldx function in conjunction with ($=) I get this error: *> :t foldr ($=)* <interactive>:1:7: Occurs check: cannot construct the infinite type: b0 = Step ao0 m0 b0 Expected type: Enumerator ao0 m0 (Step ao0 m0 b0) -> Enumeratee ao0 ao0 m0 b0 -> Enumeratee ao0 ao0 m0 b0 Actual type: Enumerator ao0 m0 (Step ao0 m0 b0) -> Enumeratee ao0 ao0 m0 b0 -> Enumerator ao0 m0 b0 In the first argument of `foldr', namely `($=)' In the expression: foldr ($=) *> :t Prelude.foldl ($=)* <interactive>:1:15: Occurs check: cannot construct the infinite type: b0 = Step ao0 m0 b0 Expected type: Enumerator ao0 m0 (Step ao0 m0 b0) -> Enumeratee ao0 ao0 m0 b0 -> Enumerator ao0 m0 (Step ao0 m0 b0) Actual type: Enumerator ao0 m0 (Step ao0 m0 b0) -> Enumeratee ao0 ao0 m0 b0 -> Enumerator ao0 m0 b0 In the first argument of `Prelude.foldl', namely `($=)' In the expression: Prelude.foldl ($=) <interactive>:1:15: Occurs check: cannot construct the infinite type: b0 = Step ao0 m0 b0 Expected type: Enumerator ao0 m0 (Step ao0 m0 b0) -> Enumeratee ao0 ao0 m0 b0 -> Enumerator ao0 m0 (Step ao0 m0 b0) Actual type: Enumerator ao0 m0 (Step ao0 m0 b0) -> Enumeratee ao0 ao0 m0 b0 -> Enumerator ao0 m0 b0 In the first argument of `Prelude.foldl', namely `($=)' In the expression: Prelude.foldl ($=) Obviously there is something I don't quite understand about the ($=) (=$) functions, how can one compose a list of enumeratees, is it even possible? Cheers. Roman.-

2 3

Re: [Haskell-cafe] Is it possible to represent such polymorphism?
by oleg＠okmij.org 04 Oct '11

04 Oct '11

sdiyazg at sjtu.edu.cn wrote: > >> generalizedFilterMap (\[x,y,z]-> if(x==1&&z==1)then [y*10] else > >> [0]) (3,1) [1,2,3,4,1,2,1,3,1,4,1,5,2] > [0,0,0,0,20,0,30,0,40,0,0] > > Of course, I could have simply used [Int] , (Num a)=>[a] or > (Int,Int,Int), but I'm trying to write code as generic as possible. As I understand, the point of generalizedFilterMap is to permit the filter function to examine several elements of the list within the current window (rather than just the current element). The step argument determines the step of sliding the window. Further, you wish to make the step argument default (setting it to 1 if not given). First of all, if you are interested in stencil computations, there are several good packages and even the whole DSLs: http://research.microsoft.com/en-us/um/people/simonpj/papers/ndp/index.htm http://arxiv.org/abs/1109.0777 http://people.csail.mit.edu/yuantang/ Second, the given code is not as generic as possible. In the above example, the filter function examines values within the window of size 3. That fact has to be specified twice: in the pattern \ [x,y,z] -> ... and as the number 3. Having to specify the same information twice is always less than satisfactory: if we expand the pattern to \ [x,y,z,u] -> we have to remember to update the other argument to the function. There are many solutions to the problem, involving as much type hacking as one may wish (some of the stencil packages above do track the size of the stencil statically, in types). Perhaps one of the simplest solution is the following > generalizedFilterMap3 :: Int -> ([a]->[a]) -> [a] -> [a] > generalizedFilterMap3 step f = concatMap f . decimate step . tails > > -- pick every n-th > decimate :: Int -> [a] -> [a] > decimate 1 lst = lst > decimate _ [] = [] > decimate n lst = head lst : decimate n (drop n lst) > > test1 = generalizedFilterMap3 1 f [1,2,3,4,1,2,1,3,1,4,1,5,2] > where > f (x:y:z:_) = if x==1&&z==1 then [y*10] else [0] > f _ = [] Again there are many, many approaches to default arguments. Perhaps the simplest, involving no typeclasses and no type computation is the following > data GMapArgs a = GMapDflt{step :: Int, > gmapf :: [a] -> [a]} > > dflt = GMapDflt{step = 1, gmapf = id} > > generalizedFilterMap dflt = generalizedFilterMap3 (step dflt) (gmapf dflt) > > test2 = generalizedFilterMap dflt{gmapf=f} [1,2,3,4,1,2,1,3,1,4,1,5,2] > where > f (x:y:z:_) = if x==1&&z==1 then [y*10] else [0] > f _ = [] > > test3 = generalizedFilterMap dflt{gmapf=f,step=2} [1,2,3,4,1,2,1,3,1,4,1,5,2] > where > f (x:y:z:_) = if x==1&&z==1 then [y*10] else [0] > f _ = [] It requires no extensions. Record puns and other Record extensions make the approach nicer. > What I want is some thing like this in C++: > > float f(char x){ return 0.1f; } > int f(double x){ return 1; } One should point out the difficulties of comparing C++ with Haskell. Haskell is designed to be higher-order, making it simple to pass functions as arguments to other functions. Your code for generic maps relies on that behavior. Try passing the overloaded C++ function 'f' defined above to some other function. That is, implement something like the following float g(??? f) { return f('x') + f(1.0); } int main(void) {printf("%g",g(f)); return 0;} One can certainly implement the above outline in C++, but it won't be simple.

1 0