Hey Timo, thanks for your warm welcome and for creating a ticket to fix this! My scenario is the following: I receive different JSON entities from an AMQP queue. I have a source to collect the events, after that I parse them into the different internal case classes and split the stream via the split function by the entity type. After that I want to transform the different entities and write the new JSON format (per entity type) to a file sink. I know I can do the transformation of the case classes to other case classes via scala code, but I thought it would be sometimes easier to do the transformations and also aggregations via the SQL or Table-API, but there are sometimes JSON arrays with objects in there and this seems to be the problem currently. I will have a look at the UDFs, maybe I can write something for my purpose.
Thanks! -- Sent from: http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/
