diff --git a/Biobase/AAseq.hs b/Biobase/AAseq.hs --- a/Biobase/AAseq.hs +++ b/Biobase/AAseq.hs @@ -5,6 +5,8 @@ {-# LANGUAGE StandaloneDeriving #-} {-# LANGUAGE PatternGuards #-} {-# LANGUAGE ViewPatterns #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE TypeFamilies #-} -- | This module has the translation tables for the genetic code. @@ -14,6 +16,7 @@ import Data.Ix (Ix(..)) import Data.Primitive.Types import Data.Tuple (swap) +import Data.Vector.Unboxed.Deriving import GHC.Base (remInt,quotInt) import qualified Data.ByteString.Char8 as BS import qualified Data.ByteString.Lazy.Char8 as BSL @@ -23,11 +26,11 @@ import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Data.Array.Repa.ExtShape -import Data.Array.Repa.Index -import Data.Array.Repa.Shape +import Data.Array.Repa.ExtShape +import Data.Array.Repa.Index +import Data.Array.Repa.Shape -import Biobase.Primary +import Biobase.Primary @@ -188,11 +191,6 @@ | Just aa <- x `lookup` charAAList = [(aa,xs)] | otherwise = [] -deriving instance Prim AA -deriving instance VGM.MVector VU.MVector AA -deriving instance VG.Vector VU.Vector AA -deriving instance VU.Unbox AA - instance (Shape sh,Show sh) => Shape (sh :. AA) where rank (sh:._) = rank sh + 1 zeroDim = zeroDim:.AA 0 @@ -250,4 +248,7 @@ instance MkAAseq T.Text where mkAAseq = VU.fromList . map toAA . T.unpack + +derivingUnbox "AA" + [t| AA -> Int |] [| unAA |] [| AA |] diff --git a/Biobase/Primary.hs b/Biobase/Primary.hs --- a/Biobase/Primary.hs +++ b/Biobase/Primary.hs @@ -1,12 +1,14 @@ -{-# LANGUAGE PackageImports #-} -{-# LANGUAGE TypeOperators #-} -{-# LANGUAGE PatternGuards #-} -{-# LANGUAGE OverlappingInstances #-} {-# LANGUAGE FlexibleInstances #-} -{-# LANGUAGE TypeSynonymInstances #-} -{-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE GeneralizedNewtypeDeriving #-} +{-# LANGUAGE MultiParamTypeClasses #-} +{-# LANGUAGE OverlappingInstances #-} +{-# LANGUAGE PackageImports #-} +{-# LANGUAGE PatternGuards #-} {-# LANGUAGE StandaloneDeriving #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE TypeFamilies #-} +{-# LANGUAGE TypeOperators #-} +{-# LANGUAGE TypeSynonymInstances #-} -- | The primary structure: interface to efficient encoding of RNA and DNA -- sequences. The design aims toward the 'vector' library and repa. In @@ -20,11 +22,12 @@ module Biobase.Primary where -import Data.Char (toUpper) -import Data.Ix (Ix(..)) -import Data.Primitive.Types -import Data.Tuple (swap) -import GHC.Base (remInt,quotInt) +import Data.Char (toUpper) +import Data.Ix (Ix(..)) +import Data.Primitive.Types +import Data.Tuple (swap) +import Data.Vector.Unboxed.Deriving +import GHC.Base (remInt,quotInt) import qualified Data.ByteString.Char8 as BS import qualified Data.ByteString.Lazy.Char8 as BSL import qualified Data.Text as T @@ -32,13 +35,14 @@ import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Data.Array.Repa.ExtShape -import Data.Array.Repa.Index -import Data.Array.Repa.Shape +import Data.Array.Repa.ExtShape +import Data.Array.Repa.Index +import Data.Array.Repa.Shape -import Biobase.Primary.Bounds +import Biobase.Primary.Bounds + -- * Convert different types of sequence representations to the internal -- "Primary Structure" representation @@ -131,12 +135,8 @@ | Just n <- x `lookup` charNucList = [(n,xs)] | otherwise = [] --- for vectors - -deriving instance Prim Nuc -deriving instance VGM.MVector VU.MVector Nuc -deriving instance VG.Vector VU.Vector Nuc -deriving instance VU.Unbox Nuc +derivingUnbox "Nuc" + [t| Nuc -> Int |] [| unNuc |] [| Nuc |] -- Shape-based indexing. Nucleotide representations go from nN (0) to nU (4), -- with additional symbols being available for specialized problems. This is a diff --git a/Biobase/Primary/Hashed.hs b/Biobase/Primary/Hashed.hs --- a/Biobase/Primary/Hashed.hs +++ b/Biobase/Primary/Hashed.hs @@ -1,30 +1,34 @@ {-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE StandaloneDeriving #-} {-# LANGUAGE GeneralizedNewtypeDeriving #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE TypeFamilies #-} -- | Fast hash functions for 'Primary' sequences. A hash is just an 'Int', so -- use these only for short sequences. +-- +-- TODO replace with standard hashing functions used by Haskell libs? module Biobase.Primary.Hashed where -import Control.Exception.Base (assert) -import Data.Ix -import Data.Primitive.Types +import Control.Exception.Base (assert) +import Data.Ix +import Data.Primitive.Types +import Data.Vector.Unboxed.Deriving import qualified Data.Vector.Generic as VG import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Biobase.Primary +import Biobase.Primary -newtype HashedPrimary = HashedPrimary Int +newtype HashedPrimary = HashedPrimary { unHashedPrimary :: Int } deriving (Eq,Ord,Ix,Read,Show,Enum,Bounded) -deriving instance Prim HashedPrimary -deriving instance VGM.MVector VU.MVector HashedPrimary -deriving instance VG.Vector VU.Vector HashedPrimary -deriving instance VU.Unbox HashedPrimary +derivingUnbox "HashedPrimary" + [t| HashedPrimary -> Int |] [| unHashedPrimary |] [| HashedPrimary |] + -- | Given a piece of primary sequence information, reduce it to an index. -- diff --git a/Biobase/Secondary.hs b/Biobase/Secondary.hs --- a/Biobase/Secondary.hs +++ b/Biobase/Secondary.hs @@ -7,6 +7,8 @@ {-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE OverlappingInstances #-} {-# LANGUAGE StandaloneDeriving #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE TypeFamilies #-} -- | Secondary structure: define basepairs as Int-tuples, the three edges, a -- nucleotide can use for pairing and the cis/trans isomerism. Both edges and @@ -17,20 +19,21 @@ module Biobase.Secondary where -import Data.Array.Repa.Index -import Data.Array.Repa.Shape -import Data.Char (toLower, toUpper) -import Data.Ix (Ix(..)) -import Data.List as L -import Data.Primitive.Types -import Data.Tuple (swap) -import GHC.Base (remInt,quotInt) +import Data.Array.Repa.Index +import Data.Array.Repa.Shape +import Data.Char (toLower, toUpper) +import Data.Ix (Ix(..)) +import Data.List as L +import Data.Primitive.Types +import Data.Tuple (swap) +import Data.Vector.Unboxed.Deriving +import GHC.Base (remInt,quotInt) import qualified Data.Vector.Generic as VG import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Biobase.Primary -import Biobase.Primary.Bounds +import Biobase.Primary +import Biobase.Primary.Bounds @@ -183,11 +186,6 @@ -- ** Instances for 'Edge' -deriving instance Prim Edge -deriving instance VGM.MVector VU.MVector Edge -deriving instance VG.Vector VU.Vector Edge -deriving instance VU.Unbox Edge - instance Bounded Edge where minBound = wc maxBound = unknownEdge @@ -204,11 +202,6 @@ -- ** Instances for 'CTisomerism' -deriving instance Prim CTisomerism -deriving instance VGM.MVector VU.MVector CTisomerism -deriving instance VG.Vector VU.Vector CTisomerism -deriving instance VU.Unbox CTisomerism - instance Bounded CTisomerism where minBound = cis maxBound = unknownCT @@ -346,4 +339,10 @@ {-# INLINE updR #-} {-# INLINE updP #-} {-# INLINE updT #-} + +derivingUnbox "Edge" + [t| Edge -> Int |] [| unEdge |] [| Edge |] + +derivingUnbox "CTisomerism" + [t| CTisomerism -> Int |] [| unCT |] [| CT |] diff --git a/Biobase/Secondary/Vienna.hs b/Biobase/Secondary/Vienna.hs --- a/Biobase/Secondary/Vienna.hs +++ b/Biobase/Secondary/Vienna.hs @@ -5,34 +5,37 @@ {-# LANGUAGE GeneralizedNewtypeDeriving #-} {-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE StandaloneDeriving #-} +{-# LANGUAGE TemplateHaskell #-} +{-# LANGUAGE TypeFamilies #-} -- | Encoding of Watson-Crick and Wobble Pairs in the Vienna RNA package style. module Biobase.Secondary.Vienna where -import Data.Array.Repa.Index -import Data.Array.Repa.Shape -import Data.Ix -import Data.Primitive.Types -import Data.Tuple (swap) -import GHC.Base (remInt,quotInt) +import Data.Array.Repa.Index +import Data.Array.Repa.Shape +import Data.Ix +import Data.Primitive.Types +import Data.Tuple (swap) +import Data.Vector.Unboxed.Deriving +import GHC.Base (remInt,quotInt) +import Prelude as P import qualified Data.Vector.Generic as VG import qualified Data.Vector.Generic.Mutable as VGM import qualified Data.Vector.Unboxed as VU -import Prelude as P -import Data.Array.Repa.ExtShape -import Data.PrimitiveArray as PA -import Data.PrimitiveArray.Zero as PA +import Data.Array.Repa.ExtShape +import Data.PrimitiveArray as PA +import Data.PrimitiveArray.Zero as PA -import Biobase.Primary -import Biobase.Primary.Bounds +import Biobase.Primary +import Biobase.Primary.Bounds -- | Use machine Ints internally -newtype ViennaPair = ViennaPair Int +newtype ViennaPair = ViennaPair { unViennaPair :: Int } deriving (Eq,Ord,Ix) instance (Shape sh,Show sh) => Shape (sh :. ViennaPair) where @@ -119,11 +122,6 @@ ] {-# NOINLINE viennaPairTable #-} -deriving instance VGM.MVector VU.MVector ViennaPair -deriving instance VG.Vector VU.Vector ViennaPair -deriving instance VU.Unbox ViennaPair -deriving instance Prim ViennaPair - instance Enum ViennaPair where toEnum x | x>=0 && x<=7 = ViennaPair x @@ -178,4 +176,7 @@ cguaP = [vpCG..vpUA] cgnsP = [vpCG..vpNS] pairToString = [(vpCG,"CG"),(vpGC,"GC"),(vpUA,"UA"),(vpAU,"AU"),(vpGU,"GU"),(vpUG,"UG"),(vpNS,"NS"),(vpNP,"NP")] + +derivingUnbox "ViennaPair" + [t| ViennaPair -> Int |] [| unViennaPair |] [| ViennaPair |] diff --git a/BiobaseXNA.cabal b/BiobaseXNA.cabal --- a/BiobaseXNA.cabal +++ b/BiobaseXNA.cabal @@ -1,9 +1,9 @@ name: BiobaseXNA -version: 0.8.2.0 +version: 0.8.3.0 author: Christian Hoener zu Siederdissen maintainer: choener@tbi.univie.ac.at homepage: http://www.tbi.univie.ac.at/~choener/ -copyright: Christian Hoener zu Siederdissen, 2011-2013 +copyright: Christian Hoener zu Siederdissen, 2011-2014 category: Bioinformatics synopsis: Efficient RNA/DNA representations license: GPL-3 @@ -40,17 +40,18 @@ library build-depends: base >3 && <5, - bytestring >= 0.10 , - containers >= 0.4 , - csv >= 0.1.2 , - file-embed >= 0.0.4.7 , - mtl >= 2.1 , - primitive >= 0.5 , - PrimitiveArray >= 0.5.4 , - repa >= 3.2 , - text >= 0.11 , - tuple >= 0.2 , - vector >= 0.10 + bytestring >= 0.10 , + containers >= 0.4 , + csv >= 0.1.2 , + file-embed >= 0.0.4.7 , + mtl >= 2.1 , + primitive >= 0.5 , + PrimitiveArray >= 0.5.4 , + repa >= 3.2 , + text >= 0.11 , + tuple >= 0.2 , + vector >= 0.10 , + vector-th-unbox >= 0.2 exposed-modules: Biobase.AAseq diff --git a/changelog b/changelog --- a/changelog +++ b/changelog @@ -1,3 +1,8 @@ +0.8.3.0 +------- + +- bugfix version: use vector-th-unbox to generate unboxed vector instances + 0.8.2.0 -------