On Thu, May 3, 2012 at 6:12 PM, yi huang <yi.codepla...@gmail.com> wrote:
> Is it intended or a bug?

> {-# LANGUAGE OverloadedStrings #-}

> import Data.ByteString.Char8 ()
> import Data.Monoid (mappend)
> import qualified Blaze.ByteString.Builder as B
> import Data.Conduit.Blaze (builderToByteString)

> import Data.Conduit
> import qualified Data.Conduit.List as CL

> main :: IO ()
> main = do
>     r <- CL.sourceList ["1", "2"]
>            $$ CL.map B.fromByteString
>            =$ CL.map (`mappend` B.flush)
>            =$ builderToByteString
>            =$ CL.consume
>     print r

> -- output: ["1","","2",""]

On Fri, May 4, 2012 at 9:59 AM, yi huang <yi.codepla...@gmail.com> wrote:
> I did some investigation to find that, `flush` is identical to
> `insertByteString S.empty`, we can make blaze-builder-conduit to ignore
> inserting of empty bytestring, that also prevents intentionally inserting of
> empty bytestring, it wouldn't cause any breakages i guess.

> On Fri, May 4, 2012 at 1:40 PM, Michael Snoyman <mich...@snoyman.com> wrote:

>> This looks buggy to me. I'll try to look at this in the next few days.

>> On Thu, May 3, 2012 at 6:12 PM, yi huang <yi.codepla...@gmail.com> wrote:
>> > Is it intended or a bug?

>> > {-# LANGUAGE OverloadedStrings #-}

>> > import Data.ByteString.Char8 ()
>> > import Data.Monoid (mappend)
>> > import qualified Blaze.ByteString.Builder as B
>> > import Data.Conduit.Blaze (builderToByteString)

>> > import Data.Conduit
>> > import qualified Data.Conduit.List as CL

>> > main :: IO ()
>> > main = do
>> >     r <- CL.sourceList ["1", "2"]
>> >            $$ CL.map B.fromByteString
>> >            =$ CL.map (`mappend` B.flush)
>> >            =$ builderToByteString
>> >            =$ CL.consume
>> >     print r

>> > -- output: ["1","","2",""]

> --
> http://yi-programmer.com/