- Libraries - Haskell.org

ObjectIO bug
by Ian Lynagh 10 Jul '08

10 Jul '08

Hi all, We've had an ObjectIO bug sitting in the GHC trac for a while, and now that ObjectIO is no longer an extralib I've closed it. As far as I know ObjectIO is not currently maintained, but if anyone is interested in picking it up then here is the bug: http://hackage.haskell.org/trac/ghc/ticket/309 Thanks Ian

1 0

HGL bugs
by Ian Lynagh 10 Jul '08

10 Jul '08

Hi all, We've had a few HGL bugs sitting in the GHC trac for a while, and now that HGL is no longer an extralib I've closed them. As far as I know HGL is not currently maintained, but if anyone is interested in picking it up then here are the bugs: http://hackage.haskell.org/trac/ghc/ticket/471 http://hackage.haskell.org/trac/ghc/ticket/742 http://hackage.haskell.org/trac/ghc/ticket/1081 http://hackage.haskell.org/trac/ghc/ticket/2286 Thanks Ian

1 0

Cabal and external dependencies
by Jim Burton 08 Jul '08

08 Jul '08

Is cabal capable of detecting missing external deps, e.g. for packages that provide bindings? Does this depend on package authors and some use of a config script seperate to .cabal? I'm just curious because I ran into this just now -- I tried to install dons' hackage-sparklines with cabal-install today (whilst reading his blog post about the same) which failed, in a way that seemed cryptic at first. I needed to notice that it was gd which failed and guess that I should install the libgd dev files. It's obvious that I would need these, you might say, but it certainly has the potential to confuse people... Jim

2 1

Cabal Help
by Dominic Steinitz 08 Jul '08

08 Jul '08

With the exception of the use of c2hs, this is almost as simple as it gets. Can someone tell me why cabal can't find IP_ICMP.chs and better still what I need to do to get cabal to build this? Thanks, Dominic. > dom@lagrange:~/networktools/ping_0> ./Setup build > Preprocessing executables for Ping-0.0... > Setup: can't find source for IP_ICMP.chs in ., dist/build/autogen > dom@lagrange:~/networktools/ping_0> more ping.cabal > Name: Ping > Version: 0.0 > License: BSD3 > Author: Dominic Steinitz > Synopsis: The Haskell equivalent of ping > > Build-type: Simple > Executable: ping > Main-Is: test.hs > Other-modules: IP_ICMP.chs > Extensions: MultiParamTypeClasses > dom@lagrange:~/networktools/ping_0> ls > _darcs dist gpl.txt IP_ICMP.chs ip_icmp.h Makefile ping.cabal ping.cabal~ Setup Setup.hi Setup.hs Setup.o test.hs

2 1

[Ticket #2393] Text.PrettyPrint.HughesPJ: Bug fixes, performance improvement
by Benedikt Huber 04 Jul '08

04 Jul '08

Hello, I'd like to propose bugfixes, documentation fixes and a performance improvement for Text.PrettyPrint.HughesPJ. The changes shouldn't effect the expected behaviour of the PP library. I've written a QuickCheck test suite for the pretty printer (to test the improvement), and found two bugs and some misconceptions/ ambiguities in the documentation. Additionally, there is a microbenchmark for the suggested improvement. Both are available at http://code.haskell.org/~bhuber/Text/ PrettyPrint/. Note that the QuickCheck tests need access to all top- level names in HughesPJ (i.e. ignore the export list). In summary, I propose to * fix a bug in fillNB and one in fillNB/sepNB * correct documentation on laws and invariants. * add more efficient implementations of vcat,hsep,hcat More specifically: (1) Bugfix fillNB: Additional case for fillNB Empty (Empty : ys) (2) Bugfix [f](cat|sep): do not allow overlapping ($$) in vertical composition (3) Lazy implementations of vcat,hcat and hsep (4) Law <t2> isn't always true (5) Invariant 5 should be made stronger (6) Change the comment about negative indentation The details follow below (maybe a little long). best regards, benedikt = = = = = = = = ======================================================================== == Details (Long) == Bug Fixes -------------------------------------- (1) Bugfix fillNB Law <l1> states that > sep (ps++[empty]++qs) = sep (ps ++ qs) > ...ditto hsep, hcat, vcat, fill... In the current implementation, this fails for the paragraph fill variants. > render' $ fill True [ text "c", text "c",empty, text "c", text "b"] > where render' = renderStyle (Style PageMode 7 1.4) >> c c c >> b The reason is a missing test for Empty in fillNB 2) Bugfix: overlap and f?(cat|sep) The specification for cat/sep: * oneLiner (hcat/hsep ps) `union` vcat ps [*] But currently cat, sep, fcat and fsep attempt to overlap the second line with the first one, i.e. they use `foldr ($$) empty ps' instead of `foldr ($+$) empty ps' [*]. I assume this is a mistake. This bug can lead to situations, where the line in the right argument of Union is actually longer: > prettyDoc$ cat [ text "a", nest 2 ( text "b") ] >> text "a"; union >> (text "b"; empty) >> (nilabove; nest 1; text "b"; empty) > renderStyle (Style PageMode 1 1) $ cat [ text "a", nest 2 ( text "b") ] >> "a b" In the implementation, we call `nilAbove False' instead of `nilAbove True' (see patch). Performance Improvements -------------------------------------- 3) Improving the space/time performance of vcat, hsep, hcat vcat (hsep,cat) is implemented in an unneccessarily strict way. We only get some output after all of vcat's arguments are evaluated and checked against being Empty. This can be improved by only checking the right argument of foldr against being Empty, and then applying an Empty-filter on the resulting Doc. Space improvement is obvious. The microbenchmark (code.haskell.org/~bhuber/Text/PrettyPrint/ HughesPJPerfCheck.hs) suggests that the improvements in time are remarkable too. Documentation fixes -------------------------------------- The QuickCheck tests revealed that 4) Law <t2> isn't always true <t2> text "" <> x = x, if x non-empty only holds if x does not start with nest (obviously, because of law <n6>). 5) The invariant 5 should be extended: A @NoDoc@ may only appear on the first line of the left argument of an union. Therefore, the right argument of an union can never be equivalent to the empty set. because this assumption is used when rendering. 6) Change the comment about negative indentation In the source code we have: -- (spaces n) generates a list of n spaces -- It should never be called with 'n' < 0, but that can happen for reasons I don't understand If we compose a <> b, and the first line of b is deeply nested, but other lines of b are not, then, because <> eats the nest, the pretty printer will try to layout some of b's lines with negative indentation: doc |0123345 ------------------ d1 |a d2 | b |c d1<>d2 |ab c| Here is the reason why this is rather unavoidable: In John Hughes paper, there is a line stating that "composing layouts does not change the layouts being composed". Another one states that "<> should eat up nests" But this leads to negative indentation - imho a user error. == End ==

7 11

Faster graph SCCs
by Iavor Diatchki 03 Jul '08

03 Jul '08

Hi, I have implemented Tarjan's algorithm for computing the strongly connected components of a graph. It is considerably faster then what's available in the "containers" package for larger graphs (see the attached picture). It would be nice to replace the existing implementation in the containers package, but I though that I'd check with the mailing list before I do this. If you want to try it out, I have put the code on hackage: http://hackage.haskell.org/cgi-bin/hackage-scripts/package/GraphSCC -Iavor

3 3

Package collections-0.3 vanished from hackage
by Benjamin Franksen 02 Jul '08

02 Jul '08

I am certain that once there was a package named collections-0.3 on the hackage server. It is gone. I managed to get the darcs version form http://code.haskell.org/collections, so this is not a show-stopper for me, but still things like that should not happen, IMO. Cheers Ben

4 3

Cabal cheers and some feature requests
by Benjamin Franksen 02 Jul '08

02 Jul '08

First: Three big cheers to Duncan for his excellent work on cabal, especially the new cabal-install. One feature I think would be uncontroversial is for cabal install to try to build haddock API docs and hscolour'ed sources. Momentarily, I still have to download (or darcs get) the package, unpack and cd into it, and then do cabal configure cabal hscolour cabal haddock cabal install --prefix=my_prefix to have the docs installed, too. Maybe add switches --with-haddock and --with-hscolour to cabal install command? (BTW: what is the trick to get the "source" links into the haddock docs? I remember there was one package I installed where this worked out of the box but I can't find it at the moment.) Another wish on my list is to be able to set defaults for cabal options in my $(HOME)/.cabal/config. Cheers Ben

3 2

GetOpt formatting improvements
by Duncan Coutts 02 Jul '08

02 Jul '08

GetOpt users, I'd like to have Cabal use the standard System.Console.GetOpt but I'd also like the output to look ok. Currently Cabal uses a private modified version of GetOpt with improved formatting and I'd like to get those improvements into the standard one and eliminate the code duplication. Here's what cabal configure looks like with the standard System.Console.GetOpt. This is what it looks like on a standard 80 column console. In particular note that the descriptions wrap around onto the following lines and look ugly. http://hpaste.org/8626 Here's a little snippet: --enable-library-for-ghci Enable compile library for use with GHCi --disable-library-for-ghci Disable compile library for use wit h GHCi --enable-split-objs Enable split library into smaller ob jects to reduce binary sizes (GHC 6.6+) --disable-split-objs Disable split library into smaller o bjects to reduce binary sizes (GHC 6.6+) --enable-executable-stripping Enable strip executables upon instal lation to reduce binary sizes --disable-executable-stripping Disable strip executables upon insta llation to reduce binary sizes --configure-option=OPT Extra option for configure --user Enable doing a per-user installation --global Disable doing a per-user installatio n --package-db=PATH Use a specific package database (to satisfy dependencies and register in) Here's the same but using the internal modified copy of GetOpt: http://hpaste.org/8626#a1 and the same snippet: --enable-library-for-ghci Enable compile library for use with GHCi --disable-library-for-ghci Disable compile library for use with GHCi --enable-split-objs Enable split library into smaller objects to reduce binary sizes (GHC 6.6+) --disable-split-objs Disable split library into smaller objects to reduce binary sizes (GHC 6.6+) --enable-executable-stripping Enable strip executables upon installation to reduce binary sizes --disable-executable-stripping Disable strip executables upon installation to reduce binary sizes --configure-option=OPT Extra option for configure --user Enable doing a per-user installation --global Disable doing a per-user installation --package-db=PATH Use a specific package database (to satisfy dependencies and register in) So there's two things to notice. One is that we wrap the description so it fits in 80 columns. The other is that we use one space rather than two of padding between the short options, the long options and the description. This gives more space for the description. Actually we go one step further and leave off the args on the short options since they're shown on the long options anyway. I'm not sure everyone wants that so I'm not really proposing it. This is what it looks like: -f FLAGS --flags=FLAGS Force values for the given... vs -f --flags=FLAGS Force values for the given flags in... Similarly: -v[n] --verbose[=n] Control verbosity (n is 0--3,... vs -v --verbose[=n] Control verbosity (n is 0--3, default... So if people think this is sensible I'll send three patches: 1. word-wrap the description to fit in 80 columns 2. use less padding between columns 3. omit the args on short options when there is also a corresponding long option though as I say, perhaps not everyone will think the 3rd change is the right thing to do. Duncan

14 24

Error accessing data files from package code in Cabal: Unkown symbol
by Alfonso Acosta 02 Jul '08

02 Jul '08

Hi! I'm working on a library package called ForSyDe. Following Cabal's documentation, I'm using the autogenerated Paths_ForSyDe module in order to access the data files of my library during execution. After installation, the Paths_ForSyDe import seems to be causing an unknown-symbol error whenever I try to compile any code which makes use of the ForSyDe library: " unknown symbol `___stginit_ForSyDezm0zi1_PathszuForSyDe_' " I'm not sure, but it seems that cabal doesn't properly link or install the Paths_ForSyDe module (I tried to include it in the cabal description file to no avail) Any suggestions about how to fix this problem? Thanks in advance, Fons P.S: I'm using Cabal-1.2.3.0

3 5