Harendra Kumar

Ahead streams might dispatch too many workers doing nothing

1

When we execute this: ``` > stream1 = Stream.sequence $ Stream.fromList [delay 2, delay 2, delay 2] > stream2 = Stream.sequence $ Stream.replicate 5 (delay 1) > Stream.fold Fold.toList $...

``` {-# INLINE chainl1 #-} chainl1 :: ParserK b IO a -> ParserK b IO (a -> a -> a) -> ParserK b IO a chainl1 p op = p...

``` if flag(dev) exposed-modules: Streamly.Internal.Data.Stream.StreamK.Alt , Streamly.Internal.Data.Stream.Type , Streamly.Internal.Data.Stream.Eliminate , Streamly.Internal.Data.Stream.Enumerate , Streamly.Internal.Data.Stream.Generate , Streamly.Internal.Data.Stream.Transform , Streamly.Internal.Data.Stream.Bottom , Streamly.Internal.Data.Stream.Exception , Streamly.Internal.Data.Stream.Expand , Streamly.Internal.Data.Stream.Lift , Streamly.Internal.Data.Stream.Reduce , Streamly.Internal.Data.Stream.Transformer , Streamly.Internal.Data.Stream.StreamDK ,...

And deprecate them in the "streamly" package. This will also remove the dependency on unicode-data package.

See the Posix/ReadDir module for posix implementations. Need to do the same for Windows. We have generic readEitherChunks in DirIO module for Windows use. The specific implementation may be faster,...

In the bench-runner cabal file.