add Acc benchmarks

I know Nikita Volkov as a Haskeller with a lot of smart logical solutions to
probems.

Acc is a tree-like structure which tree balances to the operation applied, so when a lot
of same operations applied to the structure - data tree is kept optimized for that
operation.

It is designed to be a builder-phase data type.

Therefore it lacks a lot of abilities.

Upstream thread with the author is:
nikita-volkov/acc#1

Author: Anton-Latukha

Time.hs

@@ -18,6 +18,8 @@ import qualified Data.Vector.Unboxed as UV
 import qualified Data.Vector.Storable as SV
 import qualified Data.Massiv.Array as M
 import qualified Data.RRBVector as RRB
+import qualified Acc
+import qualified GHC.Exts
 import           System.Directory
 import           System.Random
 
@@ -49,6 +51,7 @@ main = do
            , Conser "Data.Vector.Storable" sampleSVVector SV.cons
            , Conser "Data.Sequence" sampleSeq (S.<|)
            , Conser "Data.RRBVector" sampleRRB (RRB.<|)
+           , Conser "Data.Acc" sampleAcc Acc.cons
            ])
     , bgroup
         "Indexing"
@@ -72,6 +75,7 @@ main = do
            , Append "Data.Vector.Storable" sampleSVVector (SV.++) id
            , Append "Data.Sequence" sampleSeq (S.><) id
            , Append "Data.RRBVector" sampleRRB (RRB.><) id
+           , Append "Data.Acc" sampleAcc (<>) id
            ])
     , bgroup
         "Length"
@@ -84,6 +88,7 @@ main = do
            , Length "Data.Sequence" sampleSeq S.length
            , Length "Data.Massiv.Array" sampleMassivUArray M.elemsCount
            , Length "Data.RRBVector" sampleRRB length
+           , Length "Data.Acc" sampleAcc length
            ])
     , bgroup
         "Stable Sort"
@@ -116,6 +121,7 @@ main = do
            , Min "Data.Sequence" randomSampleSeq minimum
            , Min "Data.Massiv.Array" randomSampleMassivUArray M.minimum'
            , Min "Data.RRBVector" randomSampleRRB minimum
+           , Min "Data.Acc" randomSampleAcc minimum
            ])
     , bgroup
         "Max"
@@ -128,6 +134,7 @@ main = do
            , Max "Data.Sequence" randomSampleSeq maximum
            , Max "Data.Massiv.Array" randomSampleMassivUArray M.maximum'
            , Max "Data.RRBVector" randomSampleRRB maximum
+           , Max "Data.Acc" randomSampleAcc maximum
            ])
     , bgroup
         "Filter Element"
@@ -279,6 +286,9 @@ randomSampleMassivUArray i = evaluate $ force ma where
 randomSampleRRB :: Int -> IO (RRB.Vector Int)
 randomSampleRRB i = evaluate $ force $ RRB.fromList (take i (randoms (mkStdGen 0)))
 
+randomSampleAcc :: Int -> IO (Acc.Acc Int)
+randomSampleAcc i = evaluate $ force $ GHC.Exts.fromList (take i (randoms (mkStdGen 0)))
+
 sampleList :: Int -> IO [Int]
 sampleList i = evaluate $ force [1..i]
 
@@ -303,3 +313,6 @@ sampleMassivUArray i = evaluate $ force ma where
 
 sampleRRB :: Int -> IO (RRB.Vector Int)
 sampleRRB i = evaluate $ force $ RRB.fromList [1..i]
+
+sampleAcc :: Int -> IO (Acc.Acc Int)
+sampleAcc i = evaluate $ force $ GHC.Exts.fromList [1..i]

bench.cabal

@@ -23,6 +23,7 @@ benchmark time
                    , massiv
                    , rrb-vector
                    , dlist
+                   , acc
 
 executable report
   default-language: Haskell2010