July 2012 - Haskell-Cafe

lambdabot-4.2.3.3
by Dmitry Malikov 18 Jul '12

18 Jul '12

Hi. A few days ago I tried to install lambdabot package from hackage (4.2.3.2). Cabal install failed <http://hackage.haskell.org/packages/archive/lambdabot/4.2.3.2/logs/failure/…>. Then I found DanBurton's github repo <https://github.com/DanBurton/lambdabot-remedies> with some approaches to make lambdabot install fixed. All dependency packages (IOSpec, numbers) was already fixed. So I add FlexibleInstances extension to cabal file <https://github.com/dmalikov/lambdabot/commit/db7a464948d65c11b3345db524cc8f…> and upload <http://hackage.haskell.org/package/lambdabot-4.2.3.3> package to hackage. I hope that did everything right. -- Best regards, dmitry malikov !

5 8

A useful function for forking a thread, but letting the parent do setup first
by Joey Adams 18 Jul '12

18 Jul '12

Here's a useful little function: -- | Fork a thread, but wait for the main thread to perform a setup action -- using the child's 'ThreadID' before beginning work in the child thread. forkSetup :: (ThreadId -> IO (Maybe a, r)) -- ^ Setup action to be called before the thread begins working -> (a -> IO b) -- ^ What to do in the worker thread -> IO r forkSetup setup inner = mask $ \restore -> do mv <- newEmptyMVar tid <- forkIO $ join $ takeMVar mv (ma, r) <- setup tid `onException` putMVar mv (return ()) case ma of Nothing -> putMVar mv $ return () Just a -> putMVar mv $ restore $ inner a >> return () return r A question about 'mask': is it safe to use the 'restore' callback in a forked thread? Or might this be invalid in a future version of GHC? I'm aware that if forkSetup itself is called with exceptions masked, the child thread will also have exceptions masked. This seems reasonable, given that forkIO behaves the same way. Is a function like this available in some existing library? If not, where would be a good home for it? Thanks for the input!

1 0

stripSuffix
by Alvaro Gutierrez 18 Jul '12

18 Jul '12

Hi all -- Pardon me if this has been answered before: how come there's a stripPrefix in Data.List, but no matching stripSuffix? Thanks! Alvaro

4 4

What is the surefire way to handle all exceptions and make sure the program doesn't fail?
by Yifan Yu 17 Jul '12

17 Jul '12

First of all, apologise if the question is too broad. The background goes like this: I've implemented a server program in Haskell for my company intended to replace the previous one written in C which crashes a lot (and btw the technology of the company is exclusively C-based). When I chose Haskell I promised my manager (arrogantly - I actually made a bet with him), "it won't crash". Now it has been finished (with just a few hundred LOC), and my test shows that it is indeed very stable. But by looking at the code again I'm a little worried, since I'm rather new to exception handling and there're many networking-related functions in the program. I was tempted to catch (SomeException e) at the very top-level of the program and try to recursively call main to restart the server in case of any exception being thrown, but I highly doubt that is the correct and idiomatic way. There are also a number of long-running threads launched from the main thread, and exceptions thrown from these threads can't be caught by the top-level `catch' in the main thread. My main function looks like this: main :: IO () main = withSocketsDo $ do sCameraU <- socketNewPassive False 6000 sStunU <- socketNewPassive False 3478 sCmdT <- socketNewPassive True 7000 mvarCam <- newMVar M.empty mvarLog <- newMVar [] forkIO $ regCamera sCameraU mvarCam mvarLog forkIO $ updCamera mvarCam mvarLog forkIO $ stun sCameraU sStunU mvarCam mvarLog listen sCmdT 128 processCmd sCmdT mvarCam mvarLog sClose sCameraU sClose sStunU sClose sCmdT I find that I can't tell whether a function will throw any exception at all, or what exceptions will be thrown, by looking at their documentation. I can only tell if I browse the source code. So the question is, how can I determine all the exceptions that can be thrown by a given function? And what is the best way to handle situations like this, with both the long-running threads and main thread need to be restarted whenever exceptions happen. Regards, Yifan

4 6

Criterion setup/teardown functions?
by tsuraan 17 Jul '12

17 Jul '12

Is there anything in Criterion that allows for a benchmark to run some code before or after the thing that it's timing? As an example, I'd like to time a bunch of database inserts, but beforehand I want to create the target table, and afterwards I'd like to delete it. I don't really care to have the time spent on the create/delete recorded in the test run's timing, if that's possible to do.

2 2

Re: [Haskell-cafe] [haskell-cafe] Non-greedy match in Text.Regx.Posix
by Doug McIlroy 17 Jul '12

17 Jul '12

> I was exploring Text.Regex.Posix and found that I was not able to do a non-greedy match by modifying the quantifier with a "?". How can I achieve non-greedy match in Text.Regex.Posix? If it's Posix, it is greedy. Successive submatches are as long as possible subject to the constraint that a longest possible overall match be found if such exists. Doug McIlroy

1 0

Re: [Haskell-cafe] [Haskell] A riddle...
by Henning Thielemann 17 Jul '12

17 Jul '12

On Mon, 16 Jul 2012, Felipe Almeida Lessa wrote: > On Mon, Jul 16, 2012 at 12:33 PM, Vo Minh Thu <noteed(a)gmail.com> wrote: >> It seems like the infered type (and thus bounds) is different when you >> force the result to be a Color or not. Just give explicit type >> signatures and conversion functions. > > Actually, just *always* give explicit type signatures. Additionally, always compile with '-Wall' and follow the warnings. This way GHC will also warn about missing type signatures and it will even suggest type signatures. (and use haskell-cafe for riddles, please :-)

1 0

Non-greedy match in Text.Regx.Posix
by C K Kashyap 17 Jul '12

17 Jul '12

Hi all, I was exploring Text.Regex.Posix and found that I was not able to do a non-greedy match by modifying the quantifier with a "?". How can I achieve non-greedy match in Text.Regex.Posix? Regards, Kashyap

3 3

hackage compile failure with QuickCheck 2.5
by Levent Erkok 17 Jul '12

17 Jul '12

[This message is more appropriate for a hackage mailing list I presume, but that doesn't seem to exist. Let me know if there's a better place to send it.] I'm having a hackage compile failure for a newly uplodaded package that has a QuickCheck 2.5 dependence. The error message is: [13 of 13] Compiling Test.QuickCheck.All ( Test/QuickCheck/All.hs, dist/build/Test/QuickCheck/All.o ) Test/QuickCheck/All.hs:15:1: Bad interface file: /usr/local/tmp/archive/install/lib/template-haskell-2.6.0.0/ghc-7.4.1/Language/Haskell/TH.hi Something is amiss; requested module template-haskell-2.6.0.0:Language.Haskell.TH differs from name found in the interface file template-haskell:Language.Haskell.TH The full log file is at (search for "Something is a miss" in it): http://hackage.haskell.org/packages/archive/sbv/2.2/logs/failure/ghc-7.4 Needless to say, I don't see this problem when I compile this package at home with the same compiler (ghc 7.4.1) as hackage is using; also Hackage has a successfully compiled QuickCheck 2.5 package. Could it be something related to the particular cabal/ghc installation on the hackage server? In particular, I don't understand why it picks template-haskell 2.6.0.0 when there's a newer version (2.7.0.0). As far as I can see, QuickCheck doesn't put an upper limit on its template haskell version dependency. I'd appreciate any pointers with this. (Googling and questions on the #haskell irc channel didn't help much, unfortunately.) -Levent.

1 0

Haskell's type inference considered harmful [Re: [Haskell] A riddle...]
by Andreas Abel 17 Jul '12

17 Jul '12

[Is Haskell-Cafe the right place to discuss DESIGN issues of the Haskell language? This message is not about a concrete problem...] Congratulations to all that solved the riddle so quickly. (First answer in only 8 Minutes!) Now to the point of the exercise: Shocking realizations. 1. Haskell's type inference is NON-COMPOSITIONAL! In the riddle below, I am defining two things f ("rgbliste") and g ("farbliste"). Even though they are not strongly connected, but g comes after f in the definition order, the code of g influences the type of f. THAT'S WRONG! :-( It gets worse: 2. DEAD CODE changes a program's result! Yes indeed. Dead code ("farbliste") in the source cannot be removed safely. And: 3. Haskell "casts" 181935629 to Word16 value 7693 WITHOUT WARNING! Needless to say, these are serious problems for leading-edge strongly typed language. The experts are probably aware of this. But I awoke harshly from a sweet dream. Issue 3 is quite harmful and reminds me more of C than of a strongly typed language. Fixing this is not trivial. The problem is that all numeric constants have type Num a => a, and of course, type variable a can be instantiated to Int Word16 or whatever. Numeric constants need to inhabit subclasses of Num, e.g., Word16, Int and Integer /classes/ to make this behavior go away. Update: There is nothing wrong with Word16 computing modulo 2^16, but a literal like 181935629 should not get type Word16 without warning. Java asks for explicit casts in byte i = (byte) 1234876; but mindlessly also in byte i = (byte) 123; which is not a great solution either. Issues 1/2 go away with {-# LANGUAGE NoMonomorphismRestriction #-} (thx to Christian who pointed me to this). I wonder why this is not on by default?! There seems to be an unresolved dispute about this... http://hackage.haskell.org/trac/haskell-prime/wiki/NoMonomorphismRestriction But even with MonomorphismRestriction, issue 1/2 are avoidable with a more refined approach to type meta variables in the Haskell type inference: After type-checking a strongly connected component, type meta variables created for this part of the Haskell program should not be instantiated in the (topologically) later defined parts of the program. YOU DON'T WANT THE USE OF A FUNCTION INFLUENCE ITS DEFINITION! There are two alternatives what to do with remaining meta-variables of a SCC: 1. With monomorphism restriction: instantiate to the "best type". In our example below, Haskell chooses Integer. 2. Without monomorphism restriction: generalize. Choice 1 will give an error in the program below, but i rather have an error than a silent change of behavior of my program! Cheers, Andreas P.S.: In Agda, in the situation of unsolved metas in an "SCC" (a mutual block), these metas are just frozen, triggering typing errors later (since Agda cannot generalize). The user has to go back an insert more type annotations. But this is safe, whereas silent late instantiation breaks compositionality. On 16.07.2012 17:25, Andreas Abel wrote: > Today a student came to me with a piece of code that worked it executed > by itself, but produced different result in the context of his larger > problem. We cut down the example to the following: > >> import Graphics.UI.Gtk >> >> -- should produce [(26471,0,65535),... >> rgbliste = >> (map (\ i -> >> let rb = 5 * (mod (mod 181935629 (4534+i)) 100)-250+128 in >> let gb = 5 * (mod (mod 128872693 (5148+i)) 100)-250+128 in >> let bb = 5 * (mod (mod 140302469 (7578+i)) 100)-250+128 in >> let r = min 255 $ max 0 rb in >> let g = min 255 $ max 0 gb in >> let b = min 255 $ max 0 bb in >> (r*257,g*257,b*257)) [0..]) >> >> --farbliste = map (\ (r,g,b) -> Color r g b) rgbliste >> >> main :: IO () >> main = do >> print $ head rgbliste > > If you run it, it prints (26471,0,65535). > If you uncomment the def. of farbliste, it prints (44461,65535,65535). > > I was surprised. What is going on? > > Cheers, > Andreas > -- Andreas Abel <>< Du bist der geliebte Mensch. Theoretical Computer Science, University of Munich Oettingenstr. 67, D-80538 Munich, GERMANY andreas.abel(a)ifi.lmu.de http://www2.tcs.ifi.lmu.de/~abel/

3 2