Merge shortbytestring package back into bytestring wrt #444 (#471)

* Merge `shortbytestring` package back into `bytestring` wrt #444

* Fix build on ARM

Reusing compareByteArrays and avoiding
excessive pointer arithmetic.

* Speed up reverse by using byteSwap64 tricks

* Remove phase control from inlines

* Improve performance of elemIndex

* Use setByteArray in replicate

* Implement intercalate manually

* Annotate partial functions with HasCallStack

* Fix build on base < 4.12.0.0

* Add uncons/unsnoc

* Correct complexities

* Exclude reverse optimization path from ARM

It seems to cause segfaults on armv7, suggesting
there are issues with 'indexWord8ArrayAsWord64#'.

All other platforms are fine and tests pass.

* Add benchmarks for ShortByteString

* Improve inlining

* Adjust haddock identifiers

* Get rid of writeCharArray#

* Haddock fixes

* Clean up tests

* Use -fexpose-all-unfoldings

* Improve reverse

* Cleanup 'reverse'

* Fix possible GC race with foreign imports

For more information, see
  #471 (comment)

* Disable asserts in shortbytestring.c

* Remove redundant import

* Add documentation about partial functions

* Fold ShortByteString prop tests into ByteString

* Restore previous INLINEs

* Improve naming of bindings

* Consolidate error handling functions

* Remove trailing whitespace

* Fix uncons in documentation

* Rename indexWord64Array to indexWord8ArrayAsWord64

* Improve error message

* Clean up incorrect documentation

* Use div/mod instead of quot/rem

* Simplify branching in reverse

* Move asserts to Haskell

* Prefix C functions

* Fix return type of c_elem_index

* Fix documentation in unfoldrN

* Make unfoldrN more efficient

* Fix maintainer field

* Fix formatting

* Implement takeEnd, dropeEnd and splitAt manually

* Fix some haddock identifiers

* Fix unfoldrN doc

* Add a primops bounds-checking job to CI

* Document and clean up createAndTrim

* Rename errorEmptyList to errorEmptySBS

* Improve documentation for findFromEndUntil

* Improve documentation and naming

* Optimize out quotRem

* Document compareByteArraysOff

* Simplify findIndexOrLength and findFromEndUntil

* Use c_count for count

* Simplify elemIndex

* Remove use of 'mempty'

* Make sure breakSubstring is inlined into isInfixOf

* Simplify stripSuffix and stripPrefix

* Fix redundant import warnings

* Improve 'take'

* Use existing bounnds check in 'drop'

* Avoid 'create' when bytestring is empty

* Optimize filter

* Remove redundant INLINABLE

* Use shorter 'createAndTrim' in 'filter'

* Simplify 'take'

* Simplify 'drop'

* Better formatting

* Add comment to explain DNDEBUG

* Refactor elemIndex

* Optimize 'partition'

* Optimize hot loop in 'partition'

.github/workflows/ci.yml

@@ -92,3 +92,24 @@ jobs:
           ghc --version
           ghc --make -Iinclude -itests:tests/builder -o Main cbits/*.c tests/Main.hs +RTS -s
           ./Main +RTS -s
+
+  bounds-checking:
+    runs-on: ubuntu-latest
+    container:
+      image: fedora:34
+    steps:
+    - name: install deps
+      run: |
+          dnf install -y gcc gmp gmp-devel make ncurses ncurses-compat-libs xz perl
+          curl --proto '=https' --tlsv1.2 -sSf https://get-ghcup.haskell.org | BOOTSTRAP_HASKELL_NONINTERACTIVE=1 BOOTSTRAP_HASKELL_MINIMAL=1 sh
+          source ~/.ghcup/env
+          ghcup install ghc -u https://downloads.haskell.org/~ghcup/unofficial-bindists/ghc/9.3.20220124/ghc-9.3.20220124-x86_64-linux-fedora-34-bounds-checking-ddf50f4b.tar.xz --set 9.3.20220124
+          ghcup install cabal
+      shell: bash
+    - uses: actions/checkout@v1
+    - name: test
+      run: |
+          source ~/.ghcup/env
+          cabal update
+          cabal run -w ghc-9.3.20220124 --ghc-options='-fcheck-prim-bounds -fno-ignore-asserts' bytestring-tests
+      shell: bash

Changelog.md

@@ -4,7 +4,15 @@
 * [`readInt` returns `Nothing`, if the sequence of digits cannot be represented by an `Int`, instead of overflowing silently](https://github.com/haskell/bytestring/pull/309)
 * [Remove `zipWith` rewrite rule](https://github.com/haskell/bytestring/pull/387)
 
-[0.12.0.0]: https://github.com/haskell/bytestring/compare/0.11.2.0...0.12.0.0
+[0.12.0.0]: https://github.com/haskell/bytestring/compare/0.11.3.0...0.12.0.0
+
+[0.11.3.0] — Unreleased
+
+* merge `shortbytestring` package back into `bytestring` wrt [#444](https://github.com/haskell/bytestring/issues/444),
+  adding lots of additional API:
+    - [Add `all`, `any`, `append`, `break`, `breakEnd`, `breakSubstring`, `concat`, `cons`, `count`, `drop`, `dropEnd`, `dropWhile`, `dropWhileEnd`, `elem`, `elemIndex`, `elemIndices`, `filter`, `find`, `findIndex`, `findIndices`, `foldl'`, `foldl`, `foldl1'`, `foldl1`, `foldr'`, `foldr`, `foldr1'`, `foldr1`, `head`, `init`, `intercalate`, `isInfixOf`, `isPrefixOf`, `isSuffixOf`, `last`, `map`, `partition`, `replicate`, `reverse`, `singleton`, `snoc`, `span`, `spanEnd`, `split`, `splitAt`, `splitWith`, `stripPrefix`, `stripSuffix`, `tail`, `take`, `takeEnd`, `takeWhile`, `takeWhileEnd`, `uncons`, `unfoldr`, `unfoldrN`, `unsnoc`](https://github.com/haskell/bytestring/pull/471)
+
+[0.11.3.0]: https://github.com/haskell/bytestring/compare/0.11.2.0...0.11.3.0
 
 [0.11.2.0] — December 2021
 

Data/ByteString.hs

@@ -951,7 +951,7 @@ splitAt n ps@(BS x l)
     | otherwise = (BS x n, BS (plusForeignPtr x n) (l-n))
 {-# INLINE splitAt #-}
 
--- | Similar to 'P.takeWhile',
+-- | Similar to 'Prelude.takeWhile',
 -- returns the longest (possibly empty) prefix of elements
 -- satisfying the predicate.
 takeWhile :: (Word8 -> Bool) -> ByteString -> ByteString
@@ -979,7 +979,7 @@ takeWhileEnd :: (Word8 -> Bool) -> ByteString -> ByteString
 takeWhileEnd f ps = unsafeDrop (findFromEndUntil (not . f) ps) ps
 {-# INLINE takeWhileEnd #-}
 
--- | Similar to 'P.dropWhile',
+-- | Similar to 'Prelude.dropWhile',
 -- drops the longest (possibly empty) prefix of elements
 -- satisfying the predicate and returns the remainder.
 dropWhile :: (Word8 -> Bool) -> ByteString -> ByteString
@@ -997,7 +997,7 @@ dropWhile f ps = unsafeDrop (findIndexOrLength (not . f) ps) ps
     dropWhile (`eqWord8` x) = snd . spanByte x
   #-}
 
--- | Similar to 'P.dropWhileEnd',
+-- | Similar to 'Prelude.dropWhileEnd',
 -- drops the longest (possibly empty) suffix of elements
 -- satisfying the predicate and returns the remainder.
 --
@@ -1008,7 +1008,7 @@ dropWhileEnd :: (Word8 -> Bool) -> ByteString -> ByteString
 dropWhileEnd f ps = unsafeTake (findFromEndUntil (not . f) ps) ps
 {-# INLINE dropWhileEnd #-}
 
--- | Similar to 'P.break',
+-- | Similar to 'Prelude.break',
 -- returns the longest (possibly empty) prefix of elements which __do not__
 -- satisfy the predicate and the remainder of the string.
 --
@@ -1054,7 +1054,7 @@ breakByte c p = case elemIndex c p of
 breakEnd :: (Word8 -> Bool) -> ByteString -> (ByteString, ByteString)
 breakEnd  p ps = splitAt (findFromEndUntil p ps) ps
 
--- | Similar to 'P.span',
+-- | Similar to 'Prelude.span',
 -- returns the longest (possibly empty) prefix of elements
 -- satisfying the predicate and the remainder of the string.
 --

Data/ByteString/Lazy.hs

@@ -849,7 +849,7 @@ splitAt i cs0 = splitAt' i cs0
                    in (Chunk c cs', cs'')
 
 
--- | Similar to 'P.takeWhile',
+-- | Similar to 'Prelude.takeWhile',
 -- returns the longest (possibly empty) prefix of elements
 -- satisfying the predicate.
 takeWhile :: (Word8 -> Bool) -> ByteString -> ByteString
@@ -882,7 +882,7 @@ takeWhileEnd f = takeWhileEnd'
                 c' | S.length c' == S.length c -> (True, Chunk c bs)
                    | otherwise                 -> (False, fromStrict c' `append` bs)
 
--- | Similar to 'P.dropWhile',
+-- | Similar to 'Prelude.dropWhile',
 -- drops the longest (possibly empty) prefix of elements
 -- satisfying the predicate and returns the remainder.
 dropWhile :: (Word8 -> Bool) -> ByteString -> ByteString
@@ -893,7 +893,7 @@ dropWhile f = dropWhile'
             n | n < S.length c -> Chunk (S.drop n c) cs
               | otherwise      -> dropWhile' cs
 
--- | Similar to 'P.dropWhileEnd',
+-- | Similar to 'Prelude.dropWhileEnd',
 -- drops the longest (possibly empty) suffix of elements
 -- satisfying the predicate and returns the remainder.
 --
@@ -916,7 +916,7 @@ dropWhileEnd f = go []
                  x' | S.null x' -> dropEndBytes xs
                     | otherwise -> List.foldl' (flip Chunk) Empty (x' : xs)
 
--- | Similar to 'P.break',
+-- | Similar to 'Prelude.break',
 -- returns the longest (possibly empty) prefix of elements which __do not__
 -- satisfy the predicate and the remainder of the string.
 --
@@ -995,7 +995,7 @@ spanByte c (LPS ps) = case (spanByte' ps) of (a,b) -> (LPS a, LPS b)
                       | otherwise  -> (x' : [], x'' : xs)
 -}
 
--- | Similar to 'P.span',
+-- | Similar to 'Prelude.span',
 -- returns the longest (possibly empty) prefix of elements
 -- satisfying the predicate and the remainder of the string.
 --

Data/ByteString/Short.hs

@@ -2,10 +2,10 @@
 
 -- |
 -- Module      : Data.ByteString.Short
--- Copyright   : (c) Duncan Coutts 2012-2013
+-- Copyright   : (c) Duncan Coutts 2012-2013, Julian Ospald 2022
 -- License     : BSD-style
 --
--- Maintainer  : duncan@community.haskell.org
+-- Maintainer  : hasufell@posteo.de
 -- Stability   : stable
 -- Portability : ghc only
 --
@@ -67,26 +67,113 @@ module Data.ByteString.Short (
     -- small unpinned strings are allocated in the same way as normal heap
     -- allocations, rather than in a separate pinned area.
 
-    -- * Conversions
-    toShort,
-    fromShort,
+    -- * Introducing and eliminating 'ShortByteString's
+    empty,
+    singleton,
     pack,
     unpack,
+    fromShort,
+    toShort,
 
-    -- * Other operations
-    empty, null, length, index, indexMaybe, (!?),
+    -- * Basic interface
+    snoc,
+    cons,
+    append,
+    last,
+    tail,
+    uncons,
+    head,
+    init,
+    unsnoc,
+    null,
+    length,
 
-    -- ** Encoding validation
+    -- * Encoding validation
     isValidUtf8,
 
+    -- * Transforming ShortByteStrings
+    map,
+    reverse,
+    intercalate,
+
+    -- * Reducing 'ShortByteString's (folds)
+    foldl,
+    foldl',
+    foldl1,
+    foldl1',
+
+    foldr,
+    foldr',
+    foldr1,
+    foldr1',
+
+    -- ** Special folds
+    all,
+    any,
+    concat,
+
+    -- ** Generating and unfolding ByteStrings
+    replicate,
+    unfoldr,
+    unfoldrN,
+
+    -- * Substrings
+
+    -- ** Breaking strings
+    take,
+    takeEnd,
+    takeWhileEnd,
+    takeWhile,
+    drop,
+    dropEnd,
+    dropWhile,
+    dropWhileEnd,
+    breakEnd,
+    break,
+    span,
+    spanEnd,
+    splitAt,
+    split,
+    splitWith,
+    stripSuffix,
+    stripPrefix,
+
+    -- * Predicates
+    isInfixOf,
+    isPrefixOf,
+    isSuffixOf,
+
+    -- ** Search for arbitrary substrings
+    breakSubstring,
+
+    -- * Searching ShortByteStrings
+
+    -- ** Searching by equality
+    elem,
+
+    -- ** Searching with a predicate
+    find,
+    filter,
+    partition,
+
+    -- * Indexing ShortByteStrings
+    index,
+    indexMaybe,
+    (!?),
+    elemIndex,
+    elemIndices,
+    count,
+    findIndex,
+    findIndices,
+
     -- * Low level conversions
     -- ** Packing 'Foreign.C.String.CString's and pointers
     packCString,
     packCStringLen,
 
-    -- ** Using ByteStrings as 'Foreign.C.String.CString's
+    -- ** Using ShortByteStrings as 'Foreign.C.String.CString's
     useAsCString,
-    useAsCStringLen
+    useAsCStringLen,
   ) where
 
 import Data.ByteString.Short.Internal