September 2006 - Haskell-Cafe

Re[2]: [Haskell] Proposal: unification of style of function/data/type/class definitions
by Bulat Ziganshin 11 Sep '06

11 Sep '06

Hello Taral, Sunday, September 10, 2006, 9:45:33 PM, you wrote: >> data Expr t = If (Expr Bool) (Expr t) (Expr t) >> Expr Int = Lit Int >> Expr Bool | Eq t = Eq (Expr t) (Expr t) > I find this somewhat unreadable due to the implicit "t" parameter not > showing up on the left-hand side... if you mean last line, i tend to agree. i was just a little impulsive and don't mentioned that we define existential here :) it may look as Expr Bool = Eq (Expr t) (Expr t) | Eq t Brian Hulley developed here an idea about implicit "forall" declaration just before class restrictions involving this variable, so it will be equivalent to: Expr Bool = Eq (Expr t) (Expr t) | forall t . Eq t -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

3 4

HaXml and ghci unresolved symbol
by Andrea Rossato 11 Sep '06

11 Sep '06

Hello! probably it's me, but I cannot understand what I'm doing wrong. I'm trying to learn HaXml. I've never used it before and I never did xml processing in Haskell. So I'm a total newbie!! I downloaded and compiled. Everything seems fine. I'm also able to run some examples in the related directory of the source tree. Then I try doing something myself. Very simply, a document with just one element: module Xml where import Text.XML.HaXml xml = "<a>test</a>" doc = xmlParse "test" xml elemName (Document prolog _ (Elem name _ _) _) = name When I try it: *Xml> elemName doc I get a message I've never got before (I'm a newbie!) and I don't know whether the problem is in my code (this would be my first guess), in HaXml or in GHCi. And so I cannot even start understanding out to solve it: During interactive linking, GHCi couldn't find the following symbol: TextziXMLziHaXmlziParse_xmlParse_closure This may be due to you not asking GHCi to load extra object files, archives or DLLs needed by your current session. Restart GHCi, specifying the missing library using the -L/path/to/object/dir and -lmissinglibname flags, or simply by naming the relevant files on the GHCi command line. Alternatively, this link failure might indicate a bug in GHCi. If you suspect the latter, please send a bug report to: glasgow-haskell-bugs(a)haskell.org *Xml> xmlParse is defined in Text.Xml.HaXml (xmlParse :: String -> String -> Document) and it should be imported by Text.XML.HaXml. Can you give me please some directions? Thanks for your kind attention.

5 7

Nice cygwin application to be used with ghci
by Maurício 10 Sep '06

10 Sep '06

Hi, I've just uploaded a package ('rlwrap') to cygwin that I like to use with ghci. You can use it like this: rlwrap ghcii.sh and then you will use ghci as if it were readline aware (i.e., you can press up arrow to get last typed lines etc.). 'rlwrap' is very stable and I never had unexpected results while using it. Since the issue of ghci integration with terminals has been raised here sometimes, I thought some guys here would be interested (actually, I found rlwrap looking for a better way to use ghci). Best, Maurício

2 1

Re: [Haskell] Re: Proposal: unification of style of function/data/type/class definitions
by Neil Mitchell 10 Sep '06

10 Sep '06

Hi > I don't see the problem. You can read all of Bulat's > examples as "'thing being declared' 'relationship' 'value' > given that 'context'", so this one is ""sequence" has type > "[m a] -> m [a]" given that "m" is a "Monad"". So viewed > that way, they're all consistent with each other. How about name, context, value as the three elements: class <name> | <context> where <value> data <name> | <context> = <value> type <name> | <context> = <value> <name> | <context> :: <value> For class and data, you can argue that it isn't really a <value>, its <inner> or something. But type definately has a <value>. And if you want data and type consistent, you need to put them like that. And then, if you want functions consistent with types, you have to do the above. Or at least thats what my brain interprets as logical, and on the plus side it changes data/type/class which I personally don't like, but doesn't change function declarations, which I am not aware of anyone having issue with. Thanks Neil

2 1

Re: [Haskell-cafe] HaXml and ghci unresolved symbol
by Andrea Rossato 10 Sep '06

10 Sep '06

Il Sat, Sep 09, 2006 at 05:20:55PM -0400, Mark T.B. Carroll ebbe a scrivere: > FWIW I have the same problem - I can't use HaXml with ghci. > So it's not just you. (-: Indeed! So I found my first bug in ghc... And now I'll dig into ghc bug reports to see if someone is working on the problem... This is going to be hard, but it's the way you start learning something....;-) Thanks for your kind attention. Below the full error messages with the code producing them. Andrea The code below works fine with Hugs: Main> :load xml.hs Main> main prova Main> If I try to compile it with ghc I get this: [11:16:06][andrea@laptop:~/devel/haskell/xml]$ ghc --make xml.hs -package HaXml -package hxml -o prova Chasing modules from: xml.hs Compiling Main ( xml.hs, xml.o ) Linking ... /usr/local/lib/hxml-0.2/lib/libhxml.a(HaXmlAdapter.o)(.text+0x12c1): In function `spZc_dflt': : undefined reference to `TextziXMLziHaXmlziPretty_zdwelement_infó /usr/local/lib/hxml-0.2/lib/libhxml.a(HaXmlAdapter.o)(.text+0x1435): In function `spZy_dflt': : undefined reference to `TextziXMLziHaXmlziPretty_zdwelement_infó /usr/local/lib/hxml-0.2/lib/libhxml.a(HaXmlAdapter.o)(.rodata+0xa4): undefined reference to `TextziXMLziHaXmlziPretty_zdwelement_closure` /usr/local/lib/hxml-0.2/lib/libhxml.a(HaXmlAdapter.o)(.rodata+0xc4): undefined reference to `TextziXMLziHaXmlziPretty_zdwelement_closure` collect2: ld returned 1 exit status Loading HaXml and hxml in ghci will produce: [11:15:39][andrea@laptop:~/devel/haskell/xml]$ ghci -package HaXml -package hxml ___ ___ _ / _ \ /\ /\/ __(_) / /_\// /_/ / / | | GHC Interactive, version 6.4.2, for Haskell 98. / /_\\/ __ / /___| | http://www.haskell.org/ghc/ \____/\/ /_/\____/|_| Type :? for help. Loading package base-1.0 ... linking ... done. Loading package Cabal-1.1.4 ... linking ... done. Loading package haskell98-1.0 ... linking ... done. Loading package haskell-src-1.0 ... linking ... done. Loading package HaXml-1.13.1 ... linking ... done. Loading package hxml-0.2 ... linking ... ghc-6.4.2: /usr/local/lib/hxml-0.2/lib/hxml.o: unknown symbol `TextziXMLziHaXmlziTypes_AttValue_con_info` ghc-6.4.2: unable to load package `hxml-0.2' This is the code: module Main where import Text.XML.HaXml import HaXmlAdapter -- from hxml xml = "<a>prova</a>" doc = xmlParse "tmp" xml getContent (Document prolog _ (Elem name _ content) _) = content cont = getContent doc showDoc :: [Content] -> IO () showDoc [] = return () showDoc (x:xs) = do putStrLn $ showContent x showDoc xs main = showDoc cont

3 4

nondet function
by Ashley Yakeley 10 Sep '06

10 Sep '06

Is it possible to write nondet? nondet :: a -> a -> a nondet _|_ _|_ = _|_ nondet _|_ q = q nondet p _|_ = p nondet p q = p or q nondet evaluates its arguments in parallel, and returns the first one of them to evaluate. It's thus a bit different from the "par" of GPH. This isn't referentially transparent, of course, but maybe it could be written in the IO monad: nondet p q = unsafePerformIO (getnondet p q) getnondet :: a -> a -> IO a -- Ashley Yakeley, Seattle WA

4 3

Re[2]: [Haskell] Proposal: unification of style of function/data/type/class definitions
by Bulat Ziganshin 10 Sep '06

10 Sep '06

Hello Neil, Sunday, September 10, 2006, 2:21:50 PM, you wrote: >> class Monad m | Functor m, Monoid m where ... > It also makes grep'ing easier. yes, i also had problems finding class declarations in base package >> sequence :: [m a] -> m [a] | Monad m > I don't like this. In the other two instances you are moving the most > important information (the name of the thing) to the front. In this > the name is at the front, but the instances move to the end, which > isn't really where they should be. And following the function | rule > function | predicates = body > I would have said that logically, you want: > sequence | Monad m :: [m a] -> m [a] > (of course, this might present a problem for parsing...) > Note this also makes sense compared to your: > data Data | Classes = Alternatives you have omitted "patterns, or parameters shape" part: function shapes_of_parameters | guards_for_parameters = result so, in my proposal: data T a b | Eq a, Monad b = .... and function signature obey the same rule - first write shape of function type, then narrow type variables used to some classes: addEncoding :: Encoding m -> h -> m (WithEncoding m h) | Monad m, ByteStream m h Of course, that's more important is readability. it's discussible whether we need to know that m is monad and h is stream before reading shape of function type btw, the following is anyway most readable: sequence :: [Monad a] -> Monad [a] ;) -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

1 0

Re: [Haskell-cafe] HaXml and ghci unresolved symbol
by Andrea Rossato 10 Sep '06

10 Sep '06

Il Sat, Sep 09, 2006 at 04:34:04AM +0100, Marco André F. de Almeida ebbe a scrivere: > I know this will not solve your problem, but just so that you > know that with Hugs, you code works without any problems. > If you don't want/need to compile the program, I guess one interpreter > (Hugs) is as good as the other (GHCi) ;-) > > Marco Hi Marco! For the time being, that is to say, for learning xml transformation in haskell Hugs is just fine. But I'd like to start working on a project of a command line tool with hscurses that requires xml processing. I'm going to do that mostly as a learning exercise and I'm still in the preliminary phase, looking for libraries and tools. I'd like to start writing real code for the xml part, though. Thanks, Andrea

1 0

NaN, Infinity literals
by Tamas K Papp 09 Sep '06

09 Sep '06

Hi, Is there a way to use NaN and Infinity as literals, or at least to test if a value is NaN or Infinity? I tried *Main> let nan=0/0 *Main> nan NaN *Main> nan==0/0 False so "storing" the value does not work... Thanks, Tamas

6 5

Weak pointers and referential transparency???
by Brian Hulley 09 Sep '06

09 Sep '06

Hi, I have the following data structures: import System.Mem.Weak data Proxy = ... data Model = Model { _proxiesRef :: !(Ref.T [Weak Proxy]), ...} (Ref.T is just a lifted IORef) I was writing code like: createProxy :: MonadIO m => Model -> m Proxy createProxy Model{_proxiesRef = proxiesRef} = do proxies <- Ref.read proxiesRef let proxy = Proxy{ ... } weakProxy <- liftIO $ mkWeakPtr proxy Nothing Ref.write proxiesRef $! (weakProxy : proxies) return proxy where the intention is that whenever the returned proxy is no longer used, the table will no longer hold onto it (the table gets scavenged whenever I ask for the list of values in it) so there will be no space leak as proxies are created and discarded. However when I looked at the above code more closely, it struck me that Haskell gives no guarantees that the returned proxy is actually the same object that is referenced by the table, because of referential transparency. So my question is: will the above code work as expected and is this because we are allowed to assume non-referential transparency in situations like the above ie are we allowed to assume that all occurrences of the identifier "proxy" are always bound to a pointer to the same physical object and that all other uses of this value in the program are not going to cause it to be represented by multiple physical objects or otherwise split apart into registers etc? In other words, if the entry for the proxy in the table stored in the Model dies, can I be absolutely 100% sure that this means the proxy no longer exists in any shape or form in the running program? Thanks, Brian. -- Logic empowers us and Love gives us purpose. Yet still phantoms restless for eras long past, congealed in the present in unthought forms, strive mightily unseen to destroy us. http://www.metamilk.com

1 0