afb20f6 · text · git.iko.soy

git.iko.soy/text

Improve documentation, explain what is tested

commitafb20f6417ed3429f7887149a9e5e039b7b453b8 authorJasper Van der Jeugt <jaspervdj@gmail.com> dateSat 11 Jun 2011, 15:00 (15 years ago) parentcaaa43b

17 files changed+132−37

compare with master

unifiedsplit

tests/benchmarks/src/Data/Text/Benchmarks.hs+4−6

@@ -15,14 +15,13 @@ import qualified Data.Text.Benchmarks.EncodeUtf8 as EncodeUtf8

1515import qualified Data.Text.Benchmarks.Equality as Equality

1616import qualified Data.Text.Benchmarks.FileRead as FileRead

1717import qualified Data.Text.Benchmarks.FoldLines as FoldLines

18−import qualified Data.Text.Benchmarks.HtmlCombinator as HtmlCombinator

19−import qualified Data.Text.Benchmarks.Ordering as Ordering

2018import qualified Data.Text.Benchmarks.Pure as Pure

2119import qualified Data.Text.Benchmarks.ReadNumbers as ReadNumbers

2220import qualified Data.Text.Benchmarks.Replace as Replace

2321import qualified Data.Text.Benchmarks.Search as Search

2422import qualified Data.Text.Benchmarks.WordFrequencies as WordFrequencies

2523

24+import qualified Data.Text.Benchmarks.Programs.BigTable as Programs.BigTable

2625import qualified Data.Text.Benchmarks.Programs.Cut as Programs.Cut

2726import qualified Data.Text.Benchmarks.Programs.Sort as Programs.Sort

2827import qualified Data.Text.Benchmarks.Programs.StripTags as Programs.StripTags

@@ -40,12 +39,10 @@ benchmarks = do

4039 bs <- sequence

4140 [ Builder.benchmark

4241 , DecodeUtf8.benchmark (tf "russian.txt")

43− , EncodeUtf8.benchmark sink "επανάληψη 竺法蘭共譯"

42+ , EncodeUtf8.benchmark "επανάληψη 竺法蘭共譯"

4443 , Equality.benchmark (tf "japanese.txt")

4544 , FileRead.benchmark (tf "russian.txt")

4645 , FoldLines.benchmark (tf "russian.txt")

47− , HtmlCombinator.benchmark sink

48− , Ordering.benchmark (tf "russian.txt")

4946 , Pure.benchmark (tf "japanese.txt")

5047 , ReadNumbers.benchmark (tf "numbers.txt")

5148 , Replace.benchmark (tf "russian.txt") "принимая" "своем"

@@ -55,7 +52,8 @@ benchmarks = do

5552

5653 -- Program-like benchmarks

5754 ps <- bgroup "Programs" `fmap` sequence

58− [ Programs.Cut.benchmark (tf "russian.txt") sink 20 40

55+ [ Programs.BigTable.benchmark sink

56+ , Programs.Cut.benchmark (tf "russian.txt") sink 20 40

5957 , Programs.Sort.benchmark (tf "russian.txt") sink

6058 , Programs.StripTags.benchmark (tf "yiwiki.xml") sink

6159 , Programs.Throughput.benchmark (tf "russian.txt") sink

tests/benchmarks/src/Data/Text/Benchmarks/Builder.hs+4

@@ -1,5 +1,9 @@

11-- | Testing the internal builder monoid

22--

3+-- Tested in this benchmark:

4+--

5+-- * Concatenating many small strings using a builder

6+--

37{-# LANGUAGE OverloadedStrings #-}

48module Data.Text.Benchmarks.Builder

59 ( benchmark

tests/benchmarks/src/Data/Text/Benchmarks/DecodeUtf8.hs+14

@@ -1,3 +1,17 @@

1+-- | Test decoding of UTF-8

2+--

3+-- Tested in this benchmark:

4+--

5+-- * Decoding bytes using UTF-8

6+--

7+-- In some tests:

8+--

9+-- * Taking the length of the result

10+--

11+-- * Taking the init of the result

12+--

13+-- The latter are used for testing stream fusion.

14+--

115module Data.Text.Benchmarks.DecodeUtf8

216 ( benchmark

317 ) where

tests/benchmarks/src/Data/Text/Benchmarks/EncodeUtf8.hs+17−17

@@ -1,33 +1,33 @@

1+-- | UTF-8 encode a text

2+--

3+-- Tested in this benchmark:

4+--

5+-- * Replicating a string a number of times

6+--

7+-- * UTF-8 encoding it

8+--

19module Data.Text.Benchmarks.EncodeUtf8

210 ( benchmark

311 ) where

412

5−import Criterion (Benchmark, bgroup, bench)

6−import System.IO (Handle, hPutStr)

13+import Criterion (Benchmark, bgroup, bench, whnf)

714import qualified Data.ByteString as B

815import qualified Data.ByteString.Lazy as BL

916import qualified Data.Text as T

1017import qualified Data.Text.Encoding as T

11−import qualified Data.Text.IO as T

1218import qualified Data.Text.Lazy as TL

1319import qualified Data.Text.Lazy.Encoding as TL

14−import qualified Data.Text.Lazy.IO as TL

1520

16−benchmark :: Handle -> String -> IO Benchmark

17−benchmark sink string = return $ bgroup "EncodeUtf8"

18− [ bench "String" $ hPutStr sink $ concat $ replicate k string

19− , bench "Text" $ T.hPutStr sink $ T.replicate k text

20− , bench "TextLazy" $ TL.hPutStr sink $

21− TL.replicate (fromIntegral k) lazyText

22− , bench "TextByteString" $ B.hPutStr sink $

23− T.encodeUtf8 $ T.replicate k text

24− , bench "TextByteStringLazy" $ BL.hPutStr sink $

25− TL.encodeUtf8 $ TL.replicate (fromIntegral k) lazyText

26− ]

21+benchmark :: String -> IO Benchmark

22+benchmark string = do

23+ return $ bgroup "EncodeUtf8"

24+ [ bench "Text" $ whnf (B.length . T.encodeUtf8) text

25+ , bench "LazyText" $ whnf (BL.length . TL.encodeUtf8) lazyText

26+ ]

2727 where

2828 -- The string in different formats

29− text = T.pack string

30− lazyText = TL.pack string

29+ text = T.replicate k $ T.pack string

30+ lazyText = TL.replicate (fromIntegral k) $ TL.pack string

3131

3232 -- Amount

3333 k = 100000

tests/benchmarks/src/Data/Text/Benchmarks/Equality.hs+5−2

@@ -1,9 +1,12 @@

11-- | Compare a string with a copy of itself that is identical except

22-- for the last character.

33--

4+-- Tested in this benchmark:

5+--

6+-- * Comparison of strings (Eq instance)

7+--

48module Data.Text.Benchmarks.Equality

5− (

6− benchmark

9+ ( benchmark

710 ) where

811

912import Criterion (Benchmark, bgroup, bench, whnf)

tests/benchmarks/src/Data/Text/Benchmarks/FileRead.hs+4

@@ -1,5 +1,9 @@

11-- | Benchmarks simple file reading

22--

3+-- Tested in this benchmark:

4+--

5+-- * Reading a file from the disk

6+--

37module Data.Text.Benchmarks.FileRead

48 ( benchmark

59 ) where

tests/benchmarks/src/Data/Text/Benchmarks/FoldLines.hs+5−1

@@ -1,6 +1,10 @@

11-- | Read a file line-by-line using handles, and perform a fold over the lines.

22-- The fold is used here to calculate the number of lines in the file.

33--

4+-- Tested in this benchmark:

5+--

6+-- * Buffered, line-based IO

7+--

48{-# LANGUAGE BangPatterns #-}

59module Data.Text.Benchmarks.FoldLines

610 ( benchmark

@@ -14,7 +18,7 @@ import qualified Data.Text.IO as T

1418

1519benchmark :: FilePath -> IO Benchmark

1620benchmark fp = return $ bgroup "ReadLines"

17− [ bench "Text" $ withHandle $ foldLinesT (\n _ -> n + 1) (0 :: Int)

21+ [ bench "Text" $ withHandle $ foldLinesT (\n _ -> n + 1) (0 :: Int)

1822 , bench "ByteString" $ withHandle $ foldLinesB (\n _ -> n + 1) (0 :: Int)

1923 ]

2024 where

tests/benchmarks/src/Data/Text/Benchmarks/HtmlCombinator.hs → tests/benchmarks/src/Data/Text/Benchmarks/Programs/BigTable.hsrenamed+8−2

@@ -1,7 +1,13 @@

11-- | Create a large HTML table and dump it to a handle

22--

3+-- Tested in this benchmark:

4+--

5+-- * Creating a large HTML document using a builder

6+--

7+-- * Writing to a handle

8+--

39{-# LANGUAGE OverloadedStrings #-}

4−module Data.Text.Benchmarks.HtmlCombinator

10+module Data.Text.Benchmarks.Programs.BigTable

511 ( benchmark

612 ) where

713

@@ -13,7 +19,7 @@ import System.IO (Handle)

1319import qualified Data.Text as T

1420

1521benchmark :: Handle -> IO Benchmark

16−benchmark sink = return $ bench "HtmlCombinator" $ do

22+benchmark sink = return $ bench "BigTable" $ do

1723 hPutStr sink "Content-Type: text/html\n\n<table>"

1824 hPutStr sink . toLazyText . makeTable =<< rows

1925 hPutStr sink "</table>"

tests/benchmarks/src/Data/Text/Benchmarks/Programs/Cut.hs+12

@@ -1,5 +1,17 @@

11-- | Cut into a file, selecting certain columns (e.g. columns 10 to 40)

22--

3+-- Tested in this benchmark:

4+--

5+-- * Reading the file

6+--

7+-- * Splitting into lines

8+--

9+-- * Taking a number of characters from the lines

10+--

11+-- * Joining the lines

12+--

13+-- * Writing back to a handle

14+--

315module Data.Text.Benchmarks.Programs.Cut

416 ( benchmark

517 ) where

tests/benchmarks/src/Data/Text/Benchmarks/Programs/Sort.hs+12

@@ -1,5 +1,17 @@

11-- | This benchmark sorts the lines of a file, like the @sort@ unix utility.

22--

3+-- Tested in this benchmark:

4+--

5+-- * Reading the file

6+--

7+-- * Splitting into lines

8+--

9+-- * Sorting the lines

10+--

11+-- * Joining the lines

12+--

13+-- * Writing back to a handle

14+--

315{-# LANGUAGE OverloadedStrings #-}

416module Data.Text.Benchmarks.Programs.Sort

517 ( benchmark

tests/benchmarks/src/Data/Text/Benchmarks/Programs/StripTags.hs+8

@@ -2,6 +2,14 @@

22--

33-- This program was originally contributed by Petr Prokhorenkov.

44--

5+-- Tested in this benchmark:

6+--

7+-- * Reading the file

8+--

9+-- * Replacing text between HTML tags (<>) with whitespace

10+--

11+-- * Writing back to a handle

12+--

513{-# OPTIONS_GHC -fspec-constr-count=5 #-}

614module Data.Text.Benchmarks.Programs.StripTags

715 ( benchmark

tests/benchmarks/src/Data/Text/Benchmarks/Programs/Throughput.hs+8

@@ -6,6 +6,14 @@

66-- do any actual encoding/decoding here, while String and Text do have UTF-8

77-- encoding/decoding.

88--

9+-- Tested in this benchmark:

10+--

11+-- * Reading the file

12+--

13+-- * Replacing text between HTML tags (<>) with whitespace

14+--

15+-- * Writing back to a handle

16+--

917module Data.Text.Benchmarks.Programs.Throughput

1018 ( benchmark

1119 ) where

tests/benchmarks/src/Data/Text/Benchmarks/Pure.hs+4

@@ -1,5 +1,9 @@

11-- | Benchmarks various pure functions from the Text library

22--

3+-- Tested in this benchmark:

4+--

5+-- * Most pure functions defined the string types

6+--

37{-# LANGUAGE BangPatterns, GADTs, MagicHash #-}

48{-# OPTIONS_GHC -fno-warn-orphans #-}

59module Data.Text.Benchmarks.Pure

tests/benchmarks/src/Data/Text/Benchmarks/ReadNumbers.hs+11−9

@@ -12,6 +12,10 @@

1212-- The different benchmarks will only take into account the values they can

1313-- parse.

1414--

15+-- Tested in this benchmark:

16+--

17+-- * Lexing/parsing of different numerical types

18+--

1519module Data.Text.Benchmarks.ReadNumbers

1620 ( benchmark

1721 ) where

@@ -40,14 +44,14 @@ benchmark fp = do

4044 b <- B.lines `fmap` B.readFile fp

4145 bl <- BL.lines `fmap` BL.readFile fp

4246 return $ bgroup "ReadNumbers"

43− [ bench "DecimalString" $ whnf (int . string readDec) s

47+ [ bench "DecimalString" $ whnf (int . string readDec) s

4448 , bench "HexadecimalString" $ whnf (int . string readHex) s

45− , bench "DoubleString" $ whnf (double . string readFloat) s

49+ , bench "DoubleString" $ whnf (double . string readFloat) s

4650

47− , bench "DecimalText" $ whnf (int . text (T.signed T.decimal)) t

51+ , bench "DecimalText" $ whnf (int . text (T.signed T.decimal)) t

4852 , bench "HexadecimalText" $ whnf (int . text (T.signed T.hexadecimal)) t

49− , bench "DoubleText" $ whnf (double . text T.double) t

50− , bench "RationalText" $ whnf (double . text T.rational) t

53+ , bench "DoubleText" $ whnf (double . text T.double) t

54+ , bench "RationalText" $ whnf (double . text T.rational) t

5155

5256 , bench "DecimalLazyText" $

5357 whnf (int . text (TL.signed TL.decimal)) tl

@@ -58,10 +62,8 @@ benchmark fp = do

5862 , bench "RationalLazyText" $

5963 whnf (double . text TL.rational) tl

6064

61− , bench "DecimalByteString" $

62− whnf (int . byteString B.readInt) b

63− , bench "DoubleByteString" $

64− whnf (double . byteString B.readDouble) b

65+ , bench "DecimalByteString" $ whnf (int . byteString B.readInt) b

66+ , bench "DoubleByteString" $ whnf (double . byteString B.readDouble) b

6567

6668 , bench "DecimalLazyByteString" $

6769 whnf (int . byteString BL.readInt) bl

tests/benchmarks/src/Data/Text/Benchmarks/Replace.hs+4

@@ -1,5 +1,9 @@

11-- | Replace a string by another string

22--

3+-- Tested in this benchmark:

4+--

5+-- * Search and replace of a pattern in a text

6+--

37module Data.Text.Benchmarks.Replace

48 ( benchmark

59 ) where

tests/benchmarks/src/Data/Text/Benchmarks/Search.hs+4

@@ -1,5 +1,9 @@

11-- | Search for a pattern in a file, find the number of occurences

22--

3+-- Tested in this benchmark:

4+--

5+-- * Searching all occurences of a pattern using library routines

6+--

37module Data.Text.Benchmarks.Search

48 ( benchmark

59 ) where

tests/benchmarks/src/Data/Text/Benchmarks/WordFrequencies.hs+8

@@ -1,5 +1,13 @@

11-- | A word frequency count using the different string types

22--

3+-- Tested in this benchmark:

4+--

5+-- * Splitting into words

6+--

7+-- * Converting to lowercase

8+--

9+-- * Comparing: Eq/Ord instances

10+--

311module Data.Text.Benchmarks.WordFrequencies

412 ( benchmark

513 ) where