diff --git a/CHANGELOG.md b/CHANGELOG.md new file mode 100644 --- /dev/null +++ b/CHANGELOG.md @@ -0,0 +1,27 @@ +# 1.1.0 (2018-10-12) + + * Work around changes in exceptions-0.10.0. This version should + compile cleanly and work with ghc-8.6.1. + +# 1.0.4 (2018-09-02) + + * bugfix: memory leak in BGZF encoder + +# 1.0.3 (2018-08-26) + + * bugfix: seeking had been broken for a while + +# 1.0.2 (2018-07-08) + + * compatibility with ghc-8.4.1 + * bugfix: long headers were garbled when writing BAM files + * more robust imports to avoid breakage when dependencies update + +# 1.0.1 (2018-02-06) + + * Bugfix: merging of overlapping read pairs + +# 1.0.0 (2018-01-28) + + * Removed dead code and failed experiments. `biohazard` is now feature + complete by decree. diff --git a/biohazard.cabal b/biohazard.cabal --- a/biohazard.cabal +++ b/biohazard.cabal @@ -1,5 +1,5 @@ Name: biohazard -Version: 1.0.4 +Version: 1.1.0 Synopsis: bioinformatics support library Description: This is a collection of modules I separated from various bioinformatics tools. @@ -8,6 +8,7 @@ Homepage: https://bitbucket.org/ustenzel/biohazard License: BSD3 License-File: LICENSE +Extra-Source-Files: CHANGELOG.md Author: Udo Stenzel Maintainer: u.stenzel@web.de @@ -15,7 +16,7 @@ Cabal-version: >= 1.10 Build-type: Simple -Tested-with: GHC == 7.8.4, GHC == 7.10.3, GHC == 8.0.2, GHC == 8.2.1, GHC == 8.4.1, GHC == 8.4.3 +Tested-with: GHC == 7.8.4, GHC == 7.10.3, GHC == 8.0.2, GHC == 8.2.1, GHC == 8.4.3, GHC == 8.6.1 source-repository head type: git @@ -57,14 +58,14 @@ Build-depends: async >= 2.0 && < 2.3, attoparsec >= 0.10 && < 0.14, - base >= 4.7 && < 4.12, + base >= 4.7 && < 4.13, base-prelude == 1.2.0.*, bytestring >= 0.10.2 && < 0.11, - containers == 0.5.*, - exceptions >= 0.6 && < 0.9, + containers >= 0.5 && < 0.7, + exceptions >= 0.6 && < 0.11, hashable >= 1.0 && < 1.3, primitive >= 0.5 && < 0.7, - stm == 2.4.*, + stm >= 2.4 && < 2.6, text >= 1.0 && < 1.3, transformers >= 0.4.1 && < 0.6, unix >= 2.5 && < 2.8, diff --git a/src/Bio/Bam/Reader.hs b/src/Bio/Bam/Reader.hs --- a/src/Bio/Bam/Reader.hs +++ b/src/Bio/Bam/Reader.hs @@ -78,7 +78,7 @@ decodeAnyBamOrSam :: MonadIO m => BamEnumeratee m a decodeAnyBamOrSam it = isBamOrSam >>= \k -> k it -decodeAnyBamOrSamFile :: (MonadIO m, MonadMask m) +decodeAnyBamOrSamFile :: MonadBracketIO m => FilePath -> (BamMeta -> Iteratee [BamRec] m a) -> m (Iteratee [BamRec] m a) decodeAnyBamOrSamFile fn k = enumFileRandom defaultBufSize fn (decodeAnyBamOrSam k) >>= run @@ -207,13 +207,13 @@ decodeAnyBam it = do mk <- isBam ; case mk of Just k -> k it Nothing -> fail "this isn't BAM." -decodeAnyBamFile :: (MonadIO m, MonadMask m) => FilePath -> (BamMeta -> Iteratee [BamRaw] m a) -> m (Iteratee [BamRaw] m a) +decodeAnyBamFile :: MonadBracketIO m => FilePath -> (BamMeta -> Iteratee [BamRaw] m a) -> m (Iteratee [BamRaw] m a) decodeAnyBamFile fn k = enumFileRandom defaultBufSize fn (decodeAnyBam k) >>= run -concatDefaultInputs :: (MonadIO m, MonadMask m) => Enumerator' BamMeta [BamRaw] m a +concatDefaultInputs :: MonadBracketIO m => Enumerator' BamMeta [BamRaw] m a concatDefaultInputs it0 = liftIO getArgs >>= \fs -> concatInputs fs it0 -concatInputs :: (MonadIO m, MonadMask m) => [FilePath] -> Enumerator' BamMeta [BamRaw] m a +concatInputs :: MonadBracketIO m => [FilePath] -> Enumerator' BamMeta [BamRaw] m a concatInputs [ ] = enumFd defaultBufSize stdInput . decodeAnyBam >=> run concatInputs (fp0:fps) = enum1 fp0 >=> go fps where @@ -221,12 +221,12 @@ enum1 "-" = enumFd defaultBufSize stdInput . decodeAnyBam >=> run enum1 fp = enumFile defaultBufSize fp . decodeAnyBam >=> run -mergeDefaultInputs :: (MonadIO m, MonadMask m) +mergeDefaultInputs :: MonadBracketIO m => (BamMeta -> Enumeratee [BamRaw] [BamRaw] (Iteratee [BamRaw] m) a) -> Enumerator' BamMeta [BamRaw] m a mergeDefaultInputs (?) it0 = liftIO getArgs >>= \fs -> mergeInputs (?) fs it0 -mergeInputs :: (MonadIO m, MonadMask m) +mergeInputs :: MonadBracketIO m => (BamMeta -> Enumeratee [BamRaw] [BamRaw] (Iteratee [BamRaw] m) a) -> [FilePath] -> Enumerator' BamMeta [BamRaw] m a mergeInputs _ [ ] = \k -> enumFd defaultBufSize stdInput (decodeAnyBam k) >>= run diff --git a/src/Bio/Bam/Writer.hs b/src/Bio/Bam/Writer.hs --- a/src/Bio/Bam/Writer.hs +++ b/src/Bio/Bam/Writer.hs @@ -24,7 +24,6 @@ import Foreign.Marshal.Alloc ( alloca ) import System.IO ( openBinaryFile, IOMode(..) ) -import qualified Control.Monad.Catch as C import qualified Data.ByteString as B import qualified Data.ByteString.Char8 as S import qualified Data.Vector.Generic as V @@ -119,8 +118,8 @@ -- | Writes BAM encoded stuff to a file. writeBamFile :: IsBamRec r => FilePath -> BamMeta -> Iteratee [r] IO () writeBamFile fp meta = - C.bracket (liftIO $ openBinaryFile fp WriteMode) - (liftIO . hClose) + bracketIO (openBinaryFile fp WriteMode) + (hClose) (flip writeBamHandle meta) -- | write BAM encoded stuff to stdout diff --git a/src/Bio/Iteratee.hs b/src/Bio/Iteratee.hs --- a/src/Bio/Iteratee.hs +++ b/src/Bio/Iteratee.hs @@ -121,13 +121,13 @@ type Enumerator' h eo m b = (h -> Iteratee eo m b) -> m (Iteratee eo m b) type Enumeratee' h ei eo m b = (h -> Iteratee eo m b) -> Iteratee ei m (Iteratee eo m b) -enumAuxFile :: (MonadIO m, MonadMask m) => FilePath -> Iteratee S.ByteString m a -> m a +enumAuxFile :: MonadBracketIO m => FilePath -> Iteratee S.ByteString m a -> m a enumAuxFile fp it = liftIO (findAuxFile fp) >>= \f -> enumFile defaultBufSize f it >>= run -enumDefaultInputs :: (MonadIO m, MonadMask m) => Enumerator S.ByteString m a +enumDefaultInputs :: MonadBracketIO m => Enumerator S.ByteString m a enumDefaultInputs it0 = liftIO getArgs >>= flip enumInputs it0 -enumInputs :: (MonadIO m, MonadMask m) => [FilePath] -> Enumerator S.ByteString m a +enumInputs :: MonadBracketIO m => [FilePath] -> Enumerator S.ByteString m a enumInputs [] = enumFd defaultBufSize stdInput enumInputs xs = go xs where go ("-":fs) = enumFd defaultBufSize stdInput >=> go fs diff --git a/src/Bio/Iteratee/Base.hs b/src/Bio/Iteratee/Base.hs --- a/src/Bio/Iteratee/Base.hs +++ b/src/Bio/Iteratee/Base.hs @@ -169,24 +169,16 @@ instance NullPoint s => MonadTrans (Iteratee s) where - lift m = Iteratee $ \onDone _ -> m >>= flip onDone (Chunk emptyP) + lift m = Iteratee $ \onDone _ -> m >>= flip onDone (Chunk emptyP) instance (MonadIO m, Nullable s, NullPoint s) => MonadIO (Iteratee s m) where - liftIO = lift . liftIO + liftIO = lift . liftIO -instance (MonadThrow m, Nullable s, NullPoint s) => - MonadThrow (Iteratee s m) where +instance (MonadThrow m, Nullable s, NullPoint s) => MonadThrow (Iteratee s m) where throwM e = lift $ CIO.throwM e -instance (MonadCatch m, Nullable s, NullPoint s) => - MonadCatch (Iteratee s m) where +instance (MonadCatch m, Nullable s, NullPoint s) => MonadCatch (Iteratee s m) where m `catch` f = Iteratee $ \od oc -> runIter m od oc `CIO.catch` (\e -> runIter (f e) od oc) - --- prior to exceptions-0.6, these were part of MonadCatch -instance (MonadMask m, Nullable s, NullPoint s) => MonadMask (Iteratee s m) where - mask q = Iteratee $ \od oc -> CIO.mask $ \u -> runIter (q $ ilift u) od oc - uninterruptibleMask q = Iteratee $ \od oc -> CIO.uninterruptibleMask $ \u -> runIter (q $ ilift u) od oc - -- |Send 'EOF' to the @Iteratee@ and disregard the unconsumed part of the -- stream. If the iteratee is in an exception state, that exception is diff --git a/src/Bio/Iteratee/IO.hs b/src/Bio/Iteratee/IO.hs --- a/src/Bio/Iteratee/IO.hs +++ b/src/Bio/Iteratee/IO.hs @@ -15,10 +15,9 @@ where import Bio.Iteratee.Iteratee -import Bio.Prelude hiding ( bracket, loop ) -import Control.Monad.Catch ( MonadMask, bracket ) -import Control.Monad.IO.Class ( MonadIO(..) ) -import System.IO (SeekMode(..)) +import Bio.Prelude hiding ( bracket, loop ) +import Control.Monad.IO.Class ( MonadIO(..) ) +import System.IO ( SeekMode(..) ) import qualified Data.ByteString as B @@ -65,25 +64,25 @@ -enumFile' :: (MonadIO m, MonadMask m) => +enumFile' :: MonadBracketIO m => (Int -> Fd -> Enumerator s m a) -> Int -- ^Buffer size -> FilePath -> Enumerator s m a -enumFile' enumf bufsize filepath iter = bracket - (liftIO $ openFd filepath ReadOnly Nothing defaultFileFlags) - (liftIO . closeFd) +enumFile' enumf bufsize filepath iter = bracketIO + (openFd filepath ReadOnly Nothing defaultFileFlags) + (closeFd) (flip (enumf bufsize) iter) enumFile :: - (MonadIO m, MonadMask m) + MonadBracketIO m => Int -- ^Buffer size -> FilePath -> Enumerator Bytes m a enumFile = enumFile' enumFd enumFileRandom :: - (MonadIO m, MonadMask m) + MonadBracketIO m => Int -- ^Buffer size -> FilePath -> Enumerator Bytes m a diff --git a/src/Bio/Iteratee/Iteratee.hs b/src/Bio/Iteratee/Iteratee.hs --- a/src/Bio/Iteratee/Iteratee.hs +++ b/src/Bio/Iteratee/Iteratee.hs @@ -18,6 +18,7 @@ ,mBind_ ,ioBind ,ioBind_ + ,MonadBracketIO(..) -- ** Chunkwise Iteratees ,mapChunksM_ ,foldChunksM @@ -63,6 +64,7 @@ import Bio.Iteratee.Base import Bio.Prelude hiding (loop) +import Control.Monad.Catch as CIO import Control.Monad.IO.Class (MonadIO(..)) import Control.Monad.Trans.Class (MonadTrans(..)) @@ -285,6 +287,24 @@ infixl 1 `ioBind_` ioBind_ :: MonadIO m => IO a -> Iteratee s m b -> Iteratee s m b ioBind_ m b = Iteratee $ \onDone onCont -> liftIO m >> runIter b onDone onCont + +-- | Runs an 'Iteratee' in between an 'IO' action to acquire a resource +-- and one to release it. 'Iteratee' can't be an instance of +-- 'CIO.MonadMask', so 'CIO.bracket' isn't defined for it. However, if +-- we restrict the acquire/release actions to 'IO', which is the most +-- important use case anyway, we can directly implement this weaker +-- version. +class MonadIO m => MonadBracketIO m where + bracketIO :: IO a -> (a -> IO b) -> (a -> m c) -> m c + +instance MonadBracketIO IO where + {-# INLINE bracketIO #-} + bracketIO = CIO.bracket + +instance (MonadBracketIO m, Nullable s) => MonadBracketIO (Iteratee s m) where + {-# INLINE bracketIO #-} + bracketIO acquire release use = + Iteratee $ \od oc -> bracketIO acquire release (\h -> runIter (use h) od oc) -- | Convert one stream into another with the supplied mapping function. -- This function operates on whole chunks at a time, contrasting to