- Libraries - Haskell.org

ANN: unix-bytestring: Unix/Posix-specific functions for ByteStrings.
by wren ng thornton 20 Mar '11

20 Mar '11

-------------------------------------------- -- unix-bytestring 0.3.2 -------------------------------------------- The unix-bytestring package offers a full selection of Unix/Posix-specific functions for reading and writing ByteStrings to Fds. -------------------------------------------- -- The Story Behind It All -------------------------------------------- A while back I needed a ByteString-based version of the System.Posix.IO String-based API. I coded up the new versions and submitted a patch to the unix package for adding them in[1]. But apparently, noone's much pleased with the System.Posix.IO API! In the ensuing discussion folks brought up their specific complaints and offered suggestions. So now, the couple of functions have become enough to release as its own package--- with the hope that one day it'll be rolled into the unix package. That's where you come in. The current API is more than good enough for anything I've needed, so I need your feedback on what you want from a ByteString-based Unix/Posix I/O library. I've looked through Roel van Dijk's reverse dependency analysis[2] to find the folks currently using the unix package. The maintainers of the following packages may be particularly interested in unix-bytestring: * HFuse * iteratee * iteratee-mtl * liboleg HFuse has complained about the lack of bindings for pread(2) and pwrite(2), which are included in unix-bytestring. The only standing complaint I haven't addressed is one expressed in the iteratee, iteratee-mtl, and liboleg packages which would prefer a return type of (Either Errno _) instead of throwing exceptions. [1] http://article.gmane.org/gmane.comp.lang.haskell.libraries/15390 [2] http://bifunctor.homelinux.net/~roel/cgi-bin/hackage-scripts/package/unix -------------------------------------------- -- Links -------------------------------------------- Homepage: http://code.haskell.org/~wren/ Hackage: http://hackage.haskell.org/package/unix-bytestring Darcs: http://community.haskell.org/~wren/unix-bytestring Haddock (Darcs version): http://community.haskell.org/~wren/unix-bytestring/dist/doc/html/unix-bytes… -- Live well, ~wren

1 0

cabal removeLink denied on Mac
by Ulrich Vollert 20 Mar '11

20 Mar '11

Hi, I installed the Haskell-Platform 2011.2.0.0 on my Mac with Mac OS X 10.6.6...and I am happy with it ;-) After some problems with directories and files under /private/var/folders, I get an error message whenever I try to use Cabal to install new packages. For instance: ................................................ cabal install sqlite Warning: The package list for 'hackage.haskell.org' is 28 days old. Run 'cabal update' to get the latest list of available packages. Resolving dependencies... Configuring sqlite-0.5.2... Warning: 'include-dirs: include' directory does not exist. Warning: 'hs-source-dirs: .' directory does not exist. cabal: Missing dependency on a foreign library: * Missing header file: sqlite3-local.h * Missing C library: sqlite3 This problem can usually be solved by installing the system package that provides this library (you may need the "-dev" version). If the library is already installed but in a non-standard location then you can use the flags --extra-include-dirs= and --extra-lib-dirs= to specify where it is. cabal: Error: some packages failed to install: sqlite-0.5.2 failed while unpacking the package. The exception was: /var/folders/Ks/Ks-bTCXtGXaCQnMtMuXr4k+++TI/-Tmp-/sqlite-0.5.23531/sqlite-0.5.2: removeLink: permission denied (Operation not permitted) ................................................ "cabal update" works without an error message but generates the warning that the package list is old....? I assume that there is a permission problem with /private/var/folders but I dont know how to fix it. Regards Ulrich

1 0

Proposal: making inits and tails less strict
by Bas van Dijk 18 Mar '11

18 Mar '11

Hello, I would like to propose making inits and tails less strict: inits :: [a] -> [[a]] -inits [] = [[]] -inits (x:xs) = [[]] ++ map (x:) (inits xs) +inits xs = [] : case xs of + [] -> [] + x:xs -> map (x:) (inits xs) tails :: [a] -> [[a]] -tails [] = [[]] -tails xxs@(_:xs) = xxs : tails xs +tails xxs = xxs : case xxs of + [] -> [] + _:xs -> tails xs Having a lazier inits allows the elegant: nats = map length (inits nats) which loops for the current definition. This definition was due to John Tromp: http://www.mail-archive.com/haskell@haskell.org/msg21044.html Discussion deadline: 2 weeks. Regards, Bas

7 9

cabal install cabal-install with the latest HP on Mac OS X
by Yitzchak Gale 15 Mar '11

15 Mar '11

I installed the 64-bit version of the recent Haskell Platform release on Mac OS X 10.6.6. $ cabal --version cabal-install version 0.10.2 using version 1.10.1.0 of the Cabal library As always in the past, and as cabal install itself has often suggested, the first thing I did was to run 'cabal install cabal-install' to make sure I have the latest version. To my surprise, cabal install began installing an *older* version of itself, together with older versions of dependent libraries to support that. I didn't really mind, because I could always just remove the symlink to the older-versioned cabal executable when it finished, so I let it run its course to avoid leaving a mess by aborting. However, after a while, the installation process failed with the message: Downloading time-1.1.4... <command line>: cannot satisfy -package Cabal-1.10.1.0: Cabal-1.10.1.0-6289829715de7a109def1f0c286a5dc4 is unusable due to missing or recursive dependencies: process-1.0.1.5-3e412eee765d141be15796c32f22f7a3 Note that process-1.0.1.5 and Cabal-1.10.1.0 are already installed globally as part of the platform. Another copy of process-1.0.1.5 had been installed in userland during this problematic run. Cabal-1.8.0.6 was also installed in userland. Why did cabal try to downgrade itself to an older version, and why did it get so confused while trying to do so? Thanks, Yitz

3 3

FFI with ABI instead of C declaration
by Maurício CA 14 Mar '11

14 Mar '11

Hi, all, I wonder: why, in Haskell and also in other languages like ML, FFI is done with C type and code as references? Since, ultimately, compiled code goes to, say, ELF files or (in Windows) DLLs, shouldn't foreign interfaces be consistent with those? I believe this could simplify writing foreign interfaces. Say, we could write something like this: foreign haskell_name1 binary_name1 :: T1 -> T2 -> IO T1 foreign haskell_name2 binary_name2 :: T1 -> T3 -> IO T4 Our FFI tool would take a set of binary libraries where to look for binary_name[i], and match haskell_name[i] with those. T[i] would be made synonyms of types which would offer all capabilities consistent with the corresponding types in those libraries ABIs. Inconsistency, of course, would be an error (like if, in the example above, the first parameter of both functions would not be the same). Thanks for your thoughts, Maurício

2 2

Uncatchable error (fwd)
by Henning Thielemann 12 Mar '11

12 Mar '11

I like to forward this message from Haskell-Cafe: ---------- Forwarded message ---------- Date: Fri, 11 Mar 2011 17:04:16 +0100 (CET) From: Daniel Díaz <danieldiaz(a)asofilak.es> To: haskell-cafe(a)haskell.org Subject: [Haskell-cafe] Uncatchable error Hi, cafe, I'm working in a program where I use many connections with Network.HTTP. Sometimes, connections are closed while my program is reading them, and an error appears: <socket: XXX>: Data.ByteString.hGetLine: invalid argument (Bad file descriptor) All I need is to handle this error. The function 'catch' doesn't work. I guess this error comes from a call to 'error' function, or something similar. What I can do? Thanks in advance, Daniel Díaz _______________________________________________ It sounds to me that this is not a particular problem of Network.HTTP or Data.ByteString.hGetLine. It would certainly also happen for System.IO.hGetLine applied to a HTTP stream. So how can exceptions about closed HTTP connections be caught and where is it documented? The System.IO functions only mention EOF as possible exception.

1 0

Proposal: Generalize maybeNew
by Bas van Dijk 08 Mar '11

08 Mar '11

Hi, This is a tiny proposal to generalize Foreign.Marshal.Utils.maybeNew[1] from: maybeNew :: ( a -> IO (Ptr a)) -> (Maybe a -> IO (Ptr a)) to: maybeNew :: ( a -> IO (Ptr b)) -> (Maybe a -> IO (Ptr b)) so that it becomes consistent with maybeWith: maybeWith :: ( a -> (Ptr b -> IO c) -> IO c) -> (Maybe a -> (Ptr b -> IO c) -> IO c) Discussion period: 2 weeks. Bas [1] http://hackage.haskell.org/packages/archive/base/latest/doc/html/Foreign-Ma…

1 0

Proposal: add ByteString support to unix:System.Posix.IO API
by wren ng thornton 08 Mar '11

08 Mar '11

Hello all, I've written variants of the System.Posix.IO API for strict and lazy ByteStrings which are currently lingering in their own unpublished package[1]. It's silly to have a separate package for just four functions, so I'd like to see it combined into the unix package. I don't have access to create a new ticket with the patch, but it's available at [2]. Discussion time: 2 weeks. [1] darcs: http://community.haskell.org/~wren/bytestring-unix haddock: http://community.haskell.org/~wren/bytestring-unix/dist/doc/html/bytestring… [2] Darcs patch: http://community.haskell.org/~wren/bytestring-unix/bytestring-unix-0.1.0.dp… -- Live well, ~wren

6 15

darcs patch: Fix #4970 (time002 and time004 test failures on darwin)
by gwright＠antiope.com 04 Mar '11

04 Mar '11

1 patch for repository http://darcs.haskell.org/ghc-7.0/packages/old-time: Fri Mar 4 13:05:00 EST 2011 gwright(a)antiope.com * Fix #4970 (time002 and time004 test failures on darwin) This patch fixes #4970. The tv_usec field of struct timeval has type __int32_t, but getClockTime treats it as 64 bits. Because allocaBytes doesn't zero its memory, garbage can end up in the upper 32 bits. This causes a "picoseconds out of range" error in test002 and test004. The fix is to change the cast of (#peek struct timeval,tv_usec) p_timeval from "IO CTime" to "IO CInt" on darwin. The usec value is then handled correctly. I have no idea why this is only seen when the tests are run interactively but not in compiled code.

1 0

ByteString version of Text.Regex?
by Kathleen Fisher 04 Mar '11

04 Mar '11

I want a version of the Text.Regex library where the input is a character-based ByteString instead of a String. I haven't found one on hackage, although I'm not convinced I've looked exhaustively. (Every time I try to explore the regular expression libraries on Hackage, I end up in a maze of documentation, never sure precisely where I am or where I should be going.) It would be really helpful if there were a few English paragraphs somewhere explaining the high-level model provided by the various modules! I ended up taking a version of Text.Regex and making changes to produce a library that works for lazy character ByteStrings. Is that a reasonable thing to do? I'm going to make my software public soon, so I'll either need to release this modified library or swap my version out for a more standard one. Any suggestions on the best way to proceed? Kathleen

4 4