October 2012 - Haskell-Cafe

hdbc parametrized select
by s9gf4ult＠gmail.com 18 Oct '12

18 Oct '12

does hdbc have parametrized selects ? There is execute execute :: Statement -> [SqlValue] -> IO Integer and some other functions which get list of parameters but return IO Integer or IO () and there is other couple of functions fetchRow :: Statement -> IO (Maybe [SqlValue]) which return values of select but do not get parameters How can i execute parametrized select with hdbc ?

1 0

Re: [Haskell-cafe] Safe Haskell and instance coherence
by Mikhail Glushenkov 18 Oct '12

18 Oct '12

Hello David, On Thu, Oct 18, 2012 at 6:21 AM, David Mazieres expires 2013-01-15 PST <mazieres-f97tx4hr7aw7n8sqs9kpcvuwd2(a)temporary-address.scs.stanford.edu> wrote: > > What's interesting is that these examples doesn't violate the goals of > Safe Haskell--i.e., they don't violate type safety, encapsulation, or > semantic consistency as we defined it--but they are worrisome. I think that my example violates encapsulation in a similar way to how GND violates encapsulation in the MinList example from the Safe Haskell paper. IIUC, the problem is that all instances are assumed to be coherent (that's why it's an error to have two Monoid Int instances in the same module), but this assumption isn't actually enforced across module boundaries, even though it is reflected in the type system. GHC does print a warning after having encountered orphan instances, but that's all. > This is essentially the same thing as my Monoid example. You've got > two different dictionaries for the same type, and can pass either one > around depending on what module you imported. Are "dictionaries" something that is a part of Haskell semantics or an artefact of the implementation? > It's > not hard to cook up a contrived example where some sort of security > monitor hands over the keys to the kingdom if ever it encounters > duplicate items in a set. Auditing the code, that might seem fine > unless you understand the implementation of Set, which makes reasoning > about security a lot harder. Agreed. > What we really want is for the dictionary to be associated with the > data structure at the time of creation, not passed in as an argument > for each operation. But that's not even implementable without the > existential types extension, and also would require re-writing all the > containers, which is absolutely not what we want. This is what Scala does. Unfortunately, this can make some operations (e.g. set union) asymptotically less efficient (as it's now impossible to rely on the fact that both sets use the same associated dictionary). > Failing that, I guess we could disallow overlapping instances even > when they don't overlap in a single module. This is a whole-program > check similar to what type families requires, so could possibly be > implemented in a similar way. I'm really interested in the link-time check that is performed for type families. Is it described somewhere? > However, as with type families, making > it work with dynamic loading is going to be kind of hard. Essentially > there would have to be some run-time inspectable information about all > instances defined. I think that's why Rust chose to just disallow orphan instances :-) Even without dynamic loading, supporting all GHC extensions (e.g. FlexibleInstances) will probably be non-trivial. -- () ascii ribbon campaign - against html e-mail /\ www.asciiribbon.org - against proprietary attachments

1 0

Re: [Haskell-cafe] Safe Haskell and instance coherence
by Mikhail Glushenkov 17 Oct '12

17 Oct '12

Hello David, On Wed, Oct 17, 2012 at 6:02 PM, David Mazieres expires 2013-01-15 PST <mazieres-sebg4mvthwuzk9r27pqmr8jndi(a)temporary-address.scs.stanford.edu> wrote: > Can you elaborate on how this can be used to break the data structure > invariant? If in safe Haskell you import two modules that have > overlapping instances, you will not be able to use the two instances. > Modules that import only one instance will be able to use that > instance. Please take a look at the code example I provided: https://gist.github.com/3854294 I don't use overlapping instances or any other extensions besides Safe Haskell. By defining two orphan Ord instances for U I'm able to construct a value of type Set U that contains two equal elements: > test fromList [X,Y,X] This is what I meant by "breaking the data structure invariant". This shouldn't normally be possible: the documentation for Data.Set.insert says: "If the set already contains an element equal to the given value, it is replaced with the new value." Regarding your Monoid example, it will still be possible to make it work even if instance coherence is enforced by using a newtype wrapper (in fact, Data.Monoid already includes Sum and Product newtype wrappers that provide this functionality). -- () ascii ribbon campaign - against html e-mail /\ www.asciiribbon.org - against proprietary attachments

1 0

generalizing the writer monad
by Petr P 17 Oct '12

17 Oct '12

Hi, (this is a literate Haskell post.) lately I was playing with the Writer monad and it seems to me that it is too tightly coupled with monoids. Currently, MonadWriter makes the following assumptions: (1) The written value can be read again later. (2) For that to be possible it has to be monoid so that multiple (or zero) values can be combined. I fell say that this is a bit restricting. Sometimes, the written value can be lost - either used to compute something else or for example sent out using some IO action to a file, network etc. For example, I'd like to create an IO-based writer monad whose `tell` logs its argument somewhere - prints it, stores to a file etc. So what I'm suggesting is to have another type class between Monad and MonadWriter, let's say MonadTell, which only allows to write values, not to retrieve them later: > {-# LANGUAGE FlexibleInstances, MultiParamTypeClasses, FunctionalDependencies #-} > import Control.Monad > import Control.Monad.Trans > import qualified Control.Monad.Writer as W > import qualified Control.Monad.Reader as R > import Data.Monoid > > class Monad m => MonadTell w m where > tell :: w -> m () > tell w = writer ((), w) > writer :: (a, w) -> m a > writer ~(a, w) = tell w >> return a (We don't need fun.deps. here, they're needed in MonadWriter because of `listen`. IDK if it'd be still better to add fun.dep. just to eliminate typing problems?) And MonadWriter would be defined by inheriting from MonadTell: > class (MonadTell w m, Monoid w) => MonadWriter' w m | m -> w where > listen :: m a -> m (a, w) > pass :: m (a, w -> w) -> m a Now we could use MonadWriter as before, but we could also make more generic writers like: > newtype Log = Log String deriving Show > -- Prints logs to stdout. > instance MonadTell Log IO where > tell (Log s) = putStrLn s > > -- Collects the length of written logs. > instance Monad m => MonadTell Log (W.WriterT (Sum Int) m) where > tell (Log s) = W.tell (Sum $ length s) > > > main = do > let l = Log "Hello world" > tell l > print . getSum . W.execWriter $ (tell l :: W.Writer (Sum Int) ()) The same applies to MonadReader. We could make another type class between Monad and MonadReader just with `ask`: > class Monad m => MonadAsk r m | m -> r where > ask :: m r This would allow us to write instances like > instance MonadAsk Log IO where > ask = liftM Log getLine Does it make sense? Best regards, Petr Pudlak

2 1

poor performance when generating random text
by Dmitry Vyal 17 Oct '12

17 Oct '12

Hello anyone I've written a snippet which generates a file full of random strings. When compiled with -O2 on ghc-7.6, the generation speed is about 2Mb per second which is on par with interpreted php. That's the fact I find rather disappointing. Maybe I've missed something trivial? Any suggestions and explanations are welcome. :) % cat ext_sort.hs import qualified Data.Text as T import System.Random import Control.Exception import Control.Monad import System.IO import qualified Data.Text.IO as TI gen_string g = let (len, g') = randomR (50, 450) g in T.unfoldrN len rand_text (len, g') where rand_text (0,_) = Nothing rand_text (k,g) = let (c, g') = randomR ('a','z') g in Just (c, ((k-1), g')) write_corpus file = bracket (openFile file WriteMode) hClose $ \h -> do let size = 100000 sequence $ replicate size $ do g <- newStdGen let text = gen_string g TI.hPutStrLn h text main = do putStrLn "generating text corpus" write_corpus "test.txt" % cat ext_sort.prof Wed Oct 17 10:59 2012 Time and Allocation Profiling Report (Final) ext_sort +RTS -p -RTS total time = 32.56 secs (32558 ticks @ 1000 us, 1 processor) total alloc = 12,742,917,332 bytes (excludes profiling overheads) COST CENTRE MODULE %time %alloc gen_string.rand_text.(...) Main 70.7 69.8 gen_string Main 17.6 15.8 gen_string.rand_text Main 5.4 13.3 write_corpus.\ Main 4.3 0.8 individual inherited COST CENTRE MODULE no. entries %time %alloc %time %alloc MAIN MAIN 67 0 0.0 0.0 100.0 100.0 main Main 135 0 0.0 0.0 100.0 100.0 write_corpus Main 137 0 0.0 0.0 100.0 100.0 write_corpus.\ Main 138 1 4.3 0.8 100.0 100.0 write_corpus.\.text Main 140 100000 0.0 0.0 95.7 99.2 gen_string Main 141 100000 17.6 15.8 95.7 99.2 gen_string.g' Main 147 100000 0.0 0.0 0.0 0.0 gen_string.rand_text Main 144 25109743 5.4 13.3 77.5 83.2 gen_string.rand_text.g' Main 148 24909743 0.6 0.0 0.6 0.0 gen_string.rand_text.(...) Main 146 25009743 70.7 69.8 70.7 69.8 gen_string.rand_text.c Main 145 25009743 0.8 0.0 0.8 0.0 gen_string.len Main 143 100000 0.0 0.0 0.0 0.0 gen_string.(...) Main 142 100000 0.6 0.3 0.6 0.3

3 4

License of CloudHaskell code write by Haskell language.
by Chatsiri Ratana 17 Oct '12

17 Oct '12

Hello All, License of term of CloudHaskell[1]. It isn't GNU & GPL public license. I not sure to include it's to my project. Who know MPI write by haskell pure and not dependent on C/C++ wrapper such as Haskell-mpi[2]? I need to idea for implement OpenMPI standard by Haskell language. Who have an idea for implementing MPI by Haskell? [1]https://github.com/jepst/CloudHaskell/blob/master/LICENSE [2]http://hackage.haskell.org/package/haskell-mpi Thank in advance, Chatsiri Rattana. -- :--------------------------------------------------------

3 5

Is there a tool to see reduction steps?
by Daryoush Mehrtash 17 Oct '12

17 Oct '12

I have been given a piece of code that uses Tie-ing the Knot concept to label a tree of nodes in breath first manner. It seems to work fine, but I am having trouble expanding the code on my own to see the evaluation process. I like to know if there is a tools to use to see the reduction steps. > data Tree = Leaf | Node Tree Int Tree deriving Show > > label (Node ln _ rn) ((h:r):rest) = (Node lr h rr, r:r2) where > (lr, r1) = label ln rest > (rr, r2) = label rn r1 > label _ _ = (Leaf, []) > lt t = let (r, unused) = label t ([1..]:unused) > in r -- Daryoush Weblog: http://onfp.blogspot.com/

3 2

Call for Papers: Only Two Months Left - STVR Special Issue on Tests and Proofs
by Achim D. Brucker 17 Oct '12

17 Oct '12

Apologies for duplicates. ************************************************* **** Two months until the deadline for *** **** submitting a paper to the *** **** STVR special issue on Tests and Proofs *** ************************************************* CALL FOR PAPERS STVR Special Issue on Tests and Proofs http://lifc.univ-fcomte.fr/tap2012/stvr/ The Software Testing, Verification & Reliability (STVR) journal (http://www3.interscience.wiley.com/journal/13635/home) invites authors to submit papers to a Special Issue on Tests and Proofs. Background ========== The increasing use of software and the growing system complexity make focused software testing a challenging task. Recent years have seen an increasing industrial and academic interest in the use of static and dynamic analysis techniques together. Success has been reported combining different test techniques such as model-based testing, structural testing, or concolic testing with static techniques such as program slicing, dependencies analysis, model-checking, abstract interpretation, predicate abstraction, or verification. This special issue serves as a platform for researchers and practitioners to present theory, results, experience and advances in Tests and Proofs (TAP). Topics ====== This special issue focuses on all topics relevant to TAP. In particular, the topics of interest include, but are not limited to: * Program proving with the aid of testing techniques * New challenges in automated reasoning emerging from specificities of test generation * Verification and testing techniques combining proofs and tests * Generation of test data, oracles, or preambles by deductive techniques such as: theorem proving, model checking, symbolic execution, constraint logic programming, SAT and SMT solving * Model-based testing and verification * Automatic bug finding * Debugging of programs combining static and dynamic analysis * Transfer of concepts from testing to proving (e.g., coverage criteria) and from proving to testing * Formal frameworks for test and proof * Tool descriptions, experience reports and evaluation of test and proof * Case studies combining tests and proofs * Applying combination of test and proof techniques to new application domains such as validating security procotols or vulnerability detection of programs * The processes, techniques, and tools that support test and proof Submission Information ====================== The deadline for submissions is 17th December, 2012. Notification of decisions will be given by April 15th, 2013. All submissions must contain original unpublished work not being considered for publication elsewhere. Original extensions to conference papers - identifing clearly additional contributions - are also encouraged unless prohibited by copyright. Submissions will be refereed according to standard procedures for Software Testing, Verification and Reliability. Please submit your paper electronically using the Software Testing, Verification & Reliability manuscript submission site. Select "Special Issue Paper" and enter "Tests and Proofs" as title. Important Dates: ================ * Paper submission: December 17, 2012 * Notification: April 15, 2013 Guest Editors ============= * Achim D. Brucker, SAP AG, SAP Research, Germany http://www.brucker.ch/ * Wolfgang Grieskamp, Google, U.S.A. http://www.linkedin.com/in/wgrieskamp * Jacques Julliand, University of Franche-Comté, France http://lifc.univ-fcomte.fr/page_personnelle/accueil/8 -- Dr. Achim D. Brucker, SAP AG, SAP Research Vincenz-Priessnitz-Str. 1, D-76131 Karlsruhe, Phone: +49 6227 7-52595 http://www.brucker.ch

1 0

Issue building ghc on NetBSD/amd64
by lhagan 16 Oct '12

16 Oct '12

I'm attempting to build ghc on NetBSD/amd64 by following the instructions at http://hackage.haskell.org/trac/ghc/wiki/Building/Porting, but I haven't been able to get past the configure step on the target (see errors below). Where am I going wrong here? Alternatively, any ideas on where I can find pre-built binaries for this system? NetBSD/i386 would be easy, but I've had no luck finding a 64-bit version. Here's the configure error: ./configure --enable-hc-boot --build=x86_64-unknown-netbsd --host=x86_64-unknown-netbsd --target=x86_64-unknown-netbsd checking for gfind... no checking for find... /usr/bin/find checking for sort... /usr/bin/sort checking for ghc... no checking build system type... x86_64-unknown-netbsd checking host system type... x86_64-unknown-netbsd checking target system type... x86_64-unknown-netbsd HOST: x86_64-unknown-netbsd GHC build : x86_64-unknown-netbsd GHC host : x86_64-unknown-netbsd GHC target : x86_64-unknown-netbsd configure: Building in-tree ghc-pwd ./configure[3900]: : cannot execute - Is a directory configure: error: Building ghc-pwd failed

3 5

I would like a clarification about Enumerators and Iteratees, please :)
by Alfredo Di Napoli 16 Oct '12

16 Oct '12

Hi guys, I've started playing with Iteratee and Enumerators: very cool and addictive stuff. I have wrote this simple code: https://gist.github.com/3899017 In a nutshell, it gives back the number of occurences for a single char in case the argument passed from the command line is a single char, or the number of lines of the entire file. I've tried it first on a small file (~100MB) and then on a huge one (~3GB). As far as I understood Iteratee IO should be a smart way to do IO, avoiding to keep the entire file in memory; what I observe in the second case, instead, is a sort of memory leak. The memory grows and grows until the entire machine hangs. If I split the huge file in "small" chunks, memory comsumption is still high but the computation terminates and is fast. So my two questions: a) What am I missing? Should be memory used be constant? And how can I achieve this purpose? b) I'm using the package enumerator because as far as I can see is the most used. How does it compares with "iteratee"? Which of the two are more performant? Bye and thanks, A.

2 1