byteslice 0.2.8.0 → 0.2.9.0

raw patch · 6 files changed

+103/−17 lines, 6 filesdep ~primitivedep ~textPVP ok

version bump matches the API change (PVP)

Dependency ranges changed: primitive, text

API changes (from Hackage documentation)

Files

• CHANGELOG.md +5/−0
• byteslice.cabal +3/−2
• src/Data/Bytes.hs +1/−9
• src/Data/Bytes/Pure.hs +11/−0
• src/Data/Bytes/Text/Ascii.hs +41/−3
• src/Data/Bytes/Text/Utf8.hs +42/−3

@@ -1,5 +1,10 @@ # Revision history for byteslice +## 0.2.9.0 -- 2022-12-08++* Add `toShortText`, `toShortTextU`, and `toText` to `Data.Byte.Text.Ascii`.+* Add `fromShortText` and `fromText` to `Data.Bytes.Text.Utf8`.+ ## 0.2.8.0 -- 2022-11-21  * Add `Data.Bytes.replace` and `Data.Bytes.findIndices`

@@ -1,6 +1,6 @@ cabal-version: 2.4 name: byteslice-version: 0.2.8.0+version: 0.2.9.0 synopsis: Slicing managed and unmanaged memory description:   This library provides types that allow the user to talk about a slice of@@ -43,10 +43,11 @@   build-depends:     , base >=4.14 && <5     , bytestring >=0.10.8 && <0.12-    , primitive >=0.7 && <0.8+    , primitive >=0.7.4 && <0.8     , primitive-unlifted >=0.1.2 && <0.2     , primitive-addr >=0.1 && <0.2     , run-st >=0.1.1 && <0.2+    , text >=1.2.5     , text-short >=0.1.3 && <0.2     , tuples >=0.1 && <0.2     , vector >=0.12 && <0.14

@@ -167,7 +167,7 @@ import Cstrlen (cstringLength#) import Data.ByteString.Short.Internal (ShortByteString(SBS)) import Data.Bytes.Pure (length,fromByteArray,foldr,unsafeDrop)-import Data.Bytes.Pure (unsafeIndex)+import Data.Bytes.Pure (unsafeIndex,toShortByteString) import Data.Bytes.Types (Bytes(Bytes,array,offset)) import Data.Primitive (Array,ByteArray(ByteArray)) import Data.Text.Short (ShortText)@@ -615,14 +615,6 @@ all :: (Word8 -> Bool) -> Bytes -> Bool {-# inline all #-} all f = foldr (\b r -> f b && r) True---- | Convert the sliced 'Bytes' to an unsliced 'ShortByteString'. This--- reuses the array backing the sliced 'Bytes' if the slicing metadata--- implies that all of the bytes are used. Otherwise, it makes a copy.-toShortByteString :: Bytes -> ShortByteString-{-# inline toShortByteString #-}-toShortByteString !b = case Pure.toByteArray b of-  PM.ByteArray x -> SBS x  -- | Variant of 'toShortByteString' that unconditionally makes a copy of -- the array backing the sliced 'Bytes' even if the original array

@@ -38,6 +38,7 @@   , map   , mapU   , null+  , toShortByteString   ) where  import Prelude hiding (length,foldl,foldr,map,null)@@ -47,6 +48,7 @@ import Data.Bits (xor) import Data.Bytes.Types (Bytes(Bytes)) import Data.ByteString (ByteString)+import Data.ByteString.Short.Internal (ShortByteString(SBS)) import Data.Primitive (ByteArray(ByteArray),MutableByteArray,PrimArray(PrimArray)) import Data.Word (Word64,Word32,Word8) import Foreign.Ptr (Ptr,plusPtr)@@ -313,3 +315,12 @@ {-# inline unsafeHead #-} unsafeHead :: Bytes -> Word8 unsafeHead bs = unsafeIndex bs 0++-- | Convert the sliced 'Bytes' to an unsliced 'ShortByteString'. This+-- reuses the array backing the sliced 'Bytes' if the slicing metadata+-- implies that all of the bytes are used. Otherwise, it makes a copy.+toShortByteString :: Bytes -> ShortByteString+{-# inline toShortByteString #-}+toShortByteString !b = case toByteArray b of+  PM.ByteArray x -> SBS x+

@@ -1,3 +1,5 @@+{-# LANGUAGE BangPatterns #-}+{-# LANGUAGE CPP #-} {-# LANGUAGE TypeApplications #-}  -- | This module treats 'Bytes' data as holding ASCII text. Providing bytes@@ -9,17 +11,29 @@ module Data.Bytes.Text.Ascii   ( fromString   , decodeDecWord+  , toShortText+  , toShortTextU+#if MIN_VERSION_text(2,0,0)+  , toText+#endif   ) where -import Data.Bytes.Types (Bytes)+import Data.ByteString.Short.Internal (ShortByteString(SBS))+import Data.Bytes.Text.Latin1 (decodeDecWord)+import Data.Bytes.Types (Bytes(Bytes)) import Data.Char (ord)+import Data.Primitive (ByteArray)+import Data.Text (Text)+import Data.Text.Short (ShortText) import Data.Word (Word8)-import Data.Bytes.Text.Latin1 (decodeDecWord)  import qualified Data.Bytes.Pure as Bytes+import qualified Data.Primitive as PM+import qualified Data.Text.Array as A+import qualified Data.Text.Internal as I+import qualified Data.Text.Short.Unsafe as TS import qualified GHC.Exts as Exts - -- | Convert a 'String' consisting of only characters in the ASCII block -- to a byte sequence. Any character with a codepoint above @U+007F@ is -- replaced by @U+0000@.@@ -29,3 +43,27 @@   . map (\c -> let i = ord c in if i < 128 then fromIntegral @Int @Word8 i else 0)  -- TODO presumably also fromText and fromShortText++toShortText :: Bytes -> Maybe ShortText+{-# inline toShortText #-}+toShortText !b = case Bytes.foldr (\w acc -> w < 128 && acc) True b of+  True -> Just (TS.fromShortByteStringUnsafe (Bytes.toShortByteString b))+  False -> Nothing++toShortTextU :: ByteArray -> Maybe ShortText+{-# inline toShortTextU #-}+toShortTextU !b = case Bytes.foldr (\w acc -> w < 128 && acc) True (Bytes.fromByteArray b) of+  True -> Just (TS.fromShortByteStringUnsafe (case b of {PM.ByteArray x -> SBS x}))+  False -> Nothing++#if MIN_VERSION_text(2,0,0)+-- | Interpret byte sequence as ASCII codepoints.+-- Only available when building with @text-2.0@ and newer.+-- Returns 'Nothing' if any of the bytes are outside of the+-- range @0x00-0x7F@+toText :: Bytes -> Maybe Text+{-# inline toText #-}+toText !b@(Bytes (PM.ByteArray arr) off len) = case Bytes.foldr (\w acc -> w < 128 && acc) True b of+  True -> Just (I.Text (A.ByteArray arr) off len)+  False -> Nothing+#endif

@@ -1,3 +1,42 @@--- | Placeholder module in case there is demand for treating 'Bytes' as--- UTF8-encoded text-module Data.Bytes.Text.Utf8 () where+{-# language CPP #-}+{-# language BangPatterns #-}++-- | Convert 'Bytes' to and from 'Text' and 'ShortText'.+module Data.Bytes.Text.Utf8+  ( fromShortText+  , toShortText+#if MIN_VERSION_text(2,0,0)+  , fromText+#endif+  ) where++import Data.Bytes.Types (Bytes(Bytes))+import Data.Text.Short (ShortText)+import Data.Text (Text)+import Data.Primitive (ByteArray(ByteArray))++import qualified Data.Text.Short as TS+import qualified Data.Bytes as Bytes+import qualified Data.Text.Internal as I+import qualified Data.Text.Array as A++-- | Encode 'ShortText' using UTF-8. Since 'ShortText' is backed by a UTF-8+-- byte sequence, this does not perform a copy.+fromShortText :: ShortText -> Bytes+{-# inline fromShortText #-}+fromShortText = Bytes.fromShortByteString . TS.toShortByteString++-- | Attempt to interpret the byte sequence as UTF-8 encoded text. Returns+-- 'Nothing' if the bytes are not UTF-8 encoded text.+toShortText :: Bytes -> Maybe ShortText+{-# inline toShortText #-}+toShortText !b = TS.fromShortByteString (Bytes.toShortByteString b)++#if MIN_VERSION_text(2,0,0)+-- | Encode 'Text' using 'UTF-8'. Only available when building with+-- @text-2.0@ and newer. Since 'Text' is backed by a UTF-8+-- byte sequence, this does not perform a copy.+fromText :: Text -> Bytes+{-# inline fromText #-}+fromText (I.Text (A.ByteArray b) off len) = Bytes (ByteArray b) off len+#endif