- Libraries - Haskell.org

[Fwd: Re: Maps, was Re: GHC source code improvement ideas]
by Christian Maeder 08 Jan '08

08 Jan '08

-------- Original Message -------- Subject: Re: Maps, was Re: GHC source code improvement ideas Date: Fri, 04 Jan 2008 14:38:22 -0800 From: Mike Hamburg <mike(a)mikehamburg.com> To: Christian Maeder <Christian.Maeder(a)dfki.de> References: <477D987C.2090806(a)gmail.com> <477E3D69.5000307(a)gmail.com> <477E5301.8010700(a)dfki.de> On Fri, 2008-01-04 at 16:38 +0100, Christian Maeder wrote: > Simon Marlow wrote: > > Regarding Data.Map, I'd be interested in trying out AVL trees instead, > > to see if they have better performance, but then we'd have to import the > > code of course. > > Surely, we want the best maps possible for ghc and as public library > (and minimize maintenance). > > The problem is to agree on a suitable interface. I would suggest to take > (or only slightly change) Daan's interface (the current Data.Map) and > improve the underlying implementation possibly using (i.e. Adrian Hey's) > AVL trees. If anyone is off tweaking the Data.Map library, I have a simple request (which I sometimes end up implementing in hacked up versions of Data.Map). There should be some way to extract an implicitly defined interval from the map, i.e. interval :: (k -> v -> Ordering) -> Map k v -> Map k v or partitionMonotoneWithKey :: (k -> v -> Bool) -> Map k v -> (Map k v, Map k v) or similar. The important thing about these operations is that they can be made to run in O(log n) time if you have the tree structure of the map, and (so far as I know) can't if you don't. An example of when this is useful is if you have a Map (a,b) c, and wish to extract the submap for a particular a-value, or a range of a-values. Mike

2 1

new language extensions
by Duncan Coutts 07 Jan '08

07 Jan '08

All, GHC 6.8 has a lot of new names for language extensions which are not yet listed in Language.Haskell.Extension: UnicodeSyntax PatternSignatures UnliftedFFITypes LiberalTypeSynonyms TypeOperators PArr RecordWildCards RecordPuns DisambiguateRecordFields OverloadedStrings GADTs MonoPatBinds RelaxedPolyRec ExtendedDefaultRules UnboxedTuples DeriveDataTypeable ConstrainedClassMethods These are the extension names you can list in the "extensions:" field in a .cabal file and are also in the {-# LANGUAGE #-} pragma. So I propose we add all of these to the Extension enumeration data type. I'd like to keep this addition separate from the proposal to make Extension a String or newtype String, since I expect even in that case we need a central list of extensions so different implementations can agree on names when they implement the same extensions, or to ensure they pick different names for different extensions. It'd also be nice if we could document all these extensions in the Language.Haskell.Extension module so we can refer to it from the cabal documentation about the extensions field. Duncan

5 10

Generalise type of forever
by Don Stewart 07 Jan '08

07 Jan '08

It's been pointed out by a couple of people that the type of forever is overly constrained. We get a few more useful programs for free if we change the type: [generalise type of 'forever' from returning m () to m a Don Stewart <dons(a)galois.com>**20080107062656] { hunk ./Control/Monad.hs 42 - , forever -- :: (Monad m) => m a -> m () + , forever -- :: (Monad m) => m a -> m a hunk ./Control/Monad.hs 190 -forever :: (Monad m) => m a -> m () +forever :: (Monad m) => m a -> m a } In particular, fail breaks out nicely. > forever Nothing Nothing Deadline: 12th January. -- Don

2 2

bytestring 0.9.0.4
by Don Stewart 06 Jan '08

06 Jan '08

2.5 years after the first release, bytestring 0.9.0.4 is now available on hackage, http://hackage.haskell.org/cgi-bin/hackage-scripts/package/bytestring-0.9.0… Changes since the 0.9 release include: * support for bytestring literals (use -XOverloadedStrings) * make Data.ByteString.Lazy.Char8.lines lazier * fixes Handle resource leak in getContents/readFile * faster isSpace operations * performance improvements The most notable change is the instance IsString for strict and lazy bytestrings, enabling bytestrings to be written as direct string literals, without needing 'pack'. That is, the following is valid: import Data.ByteString.Char8 main = print ("abcdef" :: ByteString) (and the compile will even turn the pack into a no-op at compile time). Or, using the (soon to be released pcre-light package), bytestring regex literals can be written as: compile "a(.*)b" where compile :: ByteString -> Either String Regex With -XOverloadedStrings, the 'pack'for the regex is inferred. -- Don & Duncan

2 2

Case-insensitive formats for Data.Time.Format?
by Bjorn Buckwalter 27 Dec '07

27 Dec '07

Hello all, Is there a compelling reason to not make the formats in Data.Time.Format case-insensitive when parsing? This would apply to the months, weekdays, time zone. (I see there are already two formats %p and %P corresponding (confusingly) to AM/PM and am/pm respectively.) What got me thinking about this is that I'm being supplied with dates on the format "26 DEC" which will not parse without munging. I suspect such situations are fairly common? Also, is there a way to have the parser assume a century of 20 with the format %y? Thanks, Bjorn

2 2

getOpt return record of type a rather than [a]
by Henning Thielemann 19 Dec '07

19 Dec '07

The current version of System.Console.GetOpt.getOpt returns a list of values, where the element type has usually one constructor per option. data Flag = Verbose | Version | Input String | Output String | LibDir String What I more like to receive is a record consisting of one constructor and many fields, where optional options are of type Maybe, options with multiple occurrence are of type list. data Flag = Flag { verbose :: Bool, version :: Bool, input :: Maybe FilePath, output :: Maybe FilePath, libdir :: FilePath } Then we would need data ArgDescr a = NoArg (a -> a) ReqArg (String -> a -> a) String OptArg (String -> a -> a) String e.g. OptArg (\path flags -> flags {input = Just path}) "FILE" I feel this approach is so natural, that someone must have implemented it already.

5 7

Build Log Question
by Dominic Steinitz 17 Dec '07

17 Dec '07

I was about to update the crypto package when I noticed that there is now a "build log" on each package page. What does this mean and what should I do as package maintainer? Many thanks, Dominic. >> Step 1 - configuring Crypto-3.0.3 > > Configuring Crypto-3.0.3... > Setup.hs: At least the following dependencies are missing: > NewBinary -any > >>> Terminated due to an error. >>> Step 1 - configuring ASN1-0.0.1 > > Configuring ASN1-0.0.1... > Setup.hs: At least the following dependencies are missing: > NewBinary -any > >>> Terminated due to an error.

2 1

hierarchical module names
by Henning Thielemann 12 Dec '07

12 Dec '07

From http://haskell.org/haskellwiki/How_to_write_a_Haskell_program#Hierarchical_… there is a link to http://www.haskell.org/~simonmar/lib-hierarchy.html It looks a bit out of date (no Control.Arrow, but DSP instead of Sound). Is there a more current list?

2 2

making haddock docs with inter-package links?
by Conal Elliott 10 Dec '07

10 Dec '07

I have a bunch of libraries I'd like to re-release (via hackage & darcs) with haddock 2.0 docs. Each package has doc links within the package, to ghc libs, other packages on cabal, and to some packages with haddock docs in other places (wxhaskell, in particular). Is there any way for me to get all of these links working? And if not, is there a solution in the works? Thanks, - Conal

1 0

cabal, haddock, and --hyperlink-source
by Conal Elliott 10 Dec '07

10 Dec '07

I'm getting an error message with cabal's haddock command's --hyperlink-source option. The message (below) says "haddock --hyperlink-source requires Haddock version 0.8 or later", while "haddock --version" says that I have version 2.0. I have the latest cabal and hscolour from hackage and the latest haddock from darcs. I'm using cygwin on WinXP. bash-3.2$ ./setup haddock --hyperlink-source --verbose Creating dist\doc\html\reactive (and its parents) Preprocessing library reactive-0.0... Running hscolour for reactive-0.0... Creating dist\doc\html\reactive\src (and its parents) c:\Haskell\packages\bin\HsColour.exe -print-css -odist\doc\html\reactive\src\hscolour.css c:\Haskell\packages\bin\HsColour.exe -css -anchor -odist\doc\html\reactive\src\Data-IVar.html src\Data\IVar.hs Creating dist\doc\html\reactive\src (and its parents) c:\Haskell\packages\bin\HsColour.exe -print-css -odist\doc\html\reactive\src\hscolour.css c:\Haskell\packages\bin\HsColour.exe -css -anchor -odist\doc\html\reactive\src\Data-Fun.html src\Data\Fun.hs Creating dist\doc\html\reactive\src (and its parents) c:\Haskell\packages\bin\HsColour.exe -print-css -odist\doc\html\reactive\src\hscolour.css c:\Haskell\packages\bin\HsColour.exe -css -anchor -odist\doc\html\reactive\src\Data-Reactive.html src\Data\Reactive.hs Creating dist\build\tmp (and its parents) Creating dist\doc\html\reactive (and its parents) Preprocessing library reactive-0.0... Running Haddock for reactive-0.0... setup.exe: haddock --hyperlink-source requires Haddock version 0.8 or later bash-3.2$ haddock --version Haddock version 2.0, (c) Simon Marlow 2003; ported to the GHC-API by David Waern 2006 bash-3.2$ Ideas? - Conal

4 4