robertwb commented on issue #28664:
URL: https://github.com/apache/beam/issues/28664#issuecomment-1743404283

   Supporting the "raw" types as bytes certainly is the most flexible and makes 
sense as a first pass. I do think we'll want to support json and avro (at 
least) similar to the `_create_parser` idea we have for PubSub (I extended that 
for JSON at https://github.com/apache/beam/pull/28754 ), though for KafkaIO 
this'd be in Java of course. 
   
   I just noticed that we already have 
https://github.com/apache/beam/blob/master/sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaReadSchemaTransformProvider.java
 which is actually pretty complete. We'd probably want to add an option for the 
"raw" format that would pass the bytes through directly, as well as an option 
to add the key in if desired (similar to how the attributes can be appended as 
extra fields for PubSub). 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to