@Till: The fields are hidden inside the json string so that I have to deserialize first. Also the classes do not have so much in common. It might be possible to do it with a hierarchy of group by. I'm not sure how complicated that would be, additionally I will have to send the raw String around for longer and it might be harder to extend it later.
@Aljoscha: The problem with the split is that the channels that is split into have to have the same type. In my case the input would be string and the output would be different types of object streams. I think for now I will try to split the string stream into several sub-streams using the split operator and then do the type conversion afterwards. I will still have to distribute the logic over several operators this way but its less ugly as my current solution. cheers Martin On Thu, Sep 3, 2015 at 12:20 PM, Aljoscha Krettek <aljos...@apache.org> wrote: > Hi Martin, > maybe this is what you are looking for: > https://ci.apache.org/projects/flink/flink-docs-master/apis/streaming_guide.html#output-splitting > > Regards, > Aljoscha > > On Thu, 3 Sep 2015 at 12:02 Till Rohrmann <trohrm...@apache.org> wrote: > >> Hi Martin, >> >> could grouping be a solution to your problem? >> >> Cheers, >> Till >> >> On Thu, Sep 3, 2015 at 11:56 AM, Martin Neumann <mneum...@sics.se> wrote: >> >>> Hej, >>> >>> I have a Stream of json objects of several different types. I want to >>> split this stream into several streams each of them dealing with one type. >>> (so its not partitioning) >>> >>> The only Way I found so far is writing a bunch of filters and connect >>> them to the source directly. This way I will have a lot of duplicated logic. >>> I'm basically looking for a single operator that can have multiple >>> output streams of different types. >>> >>> cheers Martin >>> >> >>
