I've been trying to implement a protocol parser in haskell and I'm pretty new to the language, especially when it comes to monads. I've been using binary-0.5.0.2 and have described the header and all the payloads of my protocol. the messages I'd like to parse look something like the following: header + (payload A, payload B, ..) where a field in the header specifies what type of payload the message has.
I have had success in parsing the first message in the bytestring, but am at a loss with how to go about reading the next messages, discarding the bytes that were read in processing the first message.
This might be rather vague, but I'd rather get input on a generalized parser than having my ugly code altered to work in this manner.
Thanks for the help
Just use a sequence of parse operations and they will consume the input as they go along.
parseAll = do
hdr <- parseHeader
pa <- parsePayloadA
pb <- parsePayloadB
...
The second element of the tuple returned by runGetState
from Data.Binary.Get
is the remaining ByteString
. You can simply keep applying your parser until you get an error or run out of bytes.
来源:https://stackoverflow.com/questions/5517620/haskell-binary-parsing