Take a Get and an InputStream and decode a value. Consumes only as much input as necessary to decode the value. Unconsumed input will be unread. If there is an error while deserializing, a DecodeException is thrown, and unconsumed part will be unread. To simplify upstream generation, all empty ByteString will be filtered out and not passed to cereal, only EOFs/Nothing will close a cereal decoder.

Examples:

>>> import qualified System.IO.Streams as Streams
>>> getFromStream (get :: Get String) =<< Streams.fromByteString (Data.ByteString.drop 1 $ runPut $ put "encode me")
*** Exception: System.IO.Streams.Cereal: cereal decode exception: too few bytes
From:	demandInput

write a Put to an OutputStream

Convert a stream of individual encoded ByteStrings to a stream of Results. Throws a DecodeException on error.

Example:

>>> Streams.toList =<< getInputStream (get :: Get String) =<< Streams.fromList (map (runPut . put) ["foo", "bar"])
["foo","bar"]

typeclass version of getInputStream

Convert a stream of serializable objects into a stream of individual ByteStrings with a Putter, while most of the time these function are not needed, they can be used in round-trip test. Example:

>>> Streams.toList =<< getInputStream (get :: Get String) =<< encodeInputStream =<< Streams.fromList ["foo","bar"]
["foo","bar"]

typeclass version of putInputStream

create an OutputStream of serializable values from an OutputStream of bytestrings with a Putter.

typeclass version of putOutputStream