547ab19 · vty · git.iko.soy

git.iko.soy/vty

perf: Process chunks of bracketed pastes only once.

The current code is `O(n^2)` because it processes the entire input each time a new chunk arrives. This gets very expensive for large inputs. This commit changes the code to process each chunk exactly twice: once to see if it contains the end of the bracketed paste, and another time to parse the whole paste and return it as an event. The code can now process more than 1MB/sec. At this point, my tests are probably more network bound, so these timings might be off, and instead be testing my network throughput. Good enough for me anyway :). Timing: * 1MB: 1s * 4MB: 3s * 7MB: 6s * 10MB: 7s * 13MB: 10s * 15MB: 12s * 20MB: 16s

commit547ab196fc14761cdbeec11760629cf32efd2017 authoriphydf <iphydf@users.noreply.github.com> dateWed 9 Mar 2022, 23:38 (5 years ago) parent98b2a83

3 files changed+55−18

compare with master

unifiedsplit

src/Graphics/Vty/Input/Classify.hs+32−11 src/Graphics/Vty/Input/Classify/Types.hs+3 src/Graphics/Vty/Input/Loop.hs+20−7

src/Graphics/Vty/Input/Classify.hs+32−11

@@ -5,6 +5,7 @@

55module Graphics.Vty.Input.Classify

66 ( classify

77 , KClass(..)

8+ , ClassifierState(..)

89 )

910where

1011

@@ -27,6 +28,17 @@ import qualified Data.ByteString as BS

2728import qualified Data.ByteString.Char8 as BS8

2829import Data.ByteString.Char8 (ByteString)

2930

31+-- | Whether the classifier is currently processing a chunked format.

32+-- Currently, only bracketed pastes use this.

33+data ClassifierState

34+ = ClassifierStart

35+ -- ^ Not processing a chunked format.

36+ | ClassifierInChunk ByteString [ByteString]

37+ -- ^ Currently processing a chunked format. The initial chunk is in the

38+ -- first argument and a reversed remainder of the chunks is collected in

39+ -- the second argument. At the end of the processing, the chunks are

40+ -- reversed and concatenated with the final chunk.

41+

3042compile :: ClassifyMap -> ByteString -> KClass

3143compile table = cl' where

3244 -- take all prefixes and create a set of these

@@ -54,18 +66,27 @@ compile table = cl' where

5466 -- neither a prefix or a full event.

5567 [] -> Invalid

5668

57−classify :: ClassifyMap -> ByteString -> KClass

58−classify table =

59− let standardClassifier = compile table

60− in \s -> case BS.uncons s of

61− _ | bracketedPasteStarted s ->

69+classify :: ClassifyMap -> ClassifierState -> ByteString -> KClass

70+classify table = process

71+ where

72+ standardClassifier = compile table

73+

74+ process ClassifierStart s =

75+ case BS.uncons s of

76+ _ | bracketedPasteStarted s ->

77+ if bracketedPasteFinished s

78+ then parseBracketedPaste s

79+ else Chunk

80+ _ | isMouseEvent s -> classifyMouseEvent s

81+ _ | isFocusEvent s -> classifyFocusEvent s

82+ Just (c,cs) | c >= 0xC2 -> classifyUtf8 c cs

83+ _ -> standardClassifier s

84+

85+ process (ClassifierInChunk p ps) s | bracketedPasteStarted p =

6286 if bracketedPasteFinished s

63− then parseBracketedPaste s

64− else Prefix

65− _ | isMouseEvent s -> classifyMouseEvent s

66− _ | isFocusEvent s -> classifyFocusEvent s

67− Just (c,cs) | c >= 0xC2 -> classifyUtf8 c cs

68− _ -> standardClassifier s

87+ then parseBracketedPaste $ BS.concat $ p:reverse (s:ps)

88+ else Chunk

89+ process ClassifierInChunk{} _ = Invalid

6990

7091classifyUtf8 :: Word8 -> ByteString -> KClass

7192classifyUtf8 c cs =

src/Graphics/Vty/Input/Classify/Types.hs+3

@@ -19,4 +19,7 @@ data KClass

1919 | Prefix

2020 -- ^ The input characters form the prefix of a valid event character

2121 -- sequence.

22+ | Chunk

23+ -- ^ The input characters are either start of a bracketed paste chunk

24+ -- or in the middle of a bracketed paste chunk.

2225 deriving(Show, Eq)

src/Graphics/Vty/Input/Loop.hs+20−7

@@ -80,9 +80,10 @@ makeLenses ''InputBuffer

8080

8181data InputState = InputState

8282 { _unprocessedBytes :: ByteString

83+ , _classifierState :: ClassifierState

8384 , _appliedConfig :: Config

8485 , _inputBuffer :: InputBuffer

85− , _classifier :: ByteString -> KClass

86+ , _classifier :: ClassifierState -> ByteString -> KClass

8687 }

8788

8889makeLenses ''InputState

@@ -154,28 +155,40 @@ applyConfig _ _ = fail "(vty) applyConfig was not provided a complete configurat

154155parseEvent :: InputM Event

155156parseEvent = do

156157 c <- use classifier

158+ s <- use classifierState

157159 b <- use unprocessedBytes

158− case c b of

160+ case c s b of

159161 Valid e remaining -> do

160162 logMsg $ "valid parse: " ++ show e

161163 logMsg $ "remaining: " ++ show remaining

164+ classifierState .= ClassifierStart

162165 unprocessedBytes .= remaining

163166 return e

164− _ -> mzero

167+ _ -> mzero

165168

166169dropInvalid :: InputM ()

167170dropInvalid = do

168171 c <- use classifier

172+ s <- use classifierState

169173 b <- use unprocessedBytes

170− when (c b == Invalid) $ do

171− logMsg "dropping input bytes"

172− unprocessedBytes .= BS8.empty

174+ case c s b of

175+ Chunk -> do

176+ classifierState .=

177+ case s of

178+ ClassifierStart -> ClassifierInChunk b []

179+ ClassifierInChunk p bs -> ClassifierInChunk p (b:bs)

180+ unprocessedBytes .= BS8.empty

181+ Invalid -> do

182+ logMsg "dropping input bytes"

183+ classifierState .= ClassifierStart

184+ unprocessedBytes .= BS8.empty

185+ _ -> return ()

173186

174187runInputProcessorLoop :: ClassifyMap -> Input -> IO ()

175188runInputProcessorLoop classifyTable input = do

176189 let bufferSize = 1024

177190 allocaArray bufferSize $ \(bufferPtr :: Ptr Word8) -> do

178− s0 <- InputState BS8.empty

191+ s0 <- InputState BS8.empty ClassifierStart

179192 <$> readIORef (_configRef input)

180193 <*> pure (InputBuffer bufferPtr bufferSize)

181194 <*> pure (classify classifyTable)