Yes, I'm aware of that. It would also be required for every split unless we cache the BOM per file. OTOH, if we do it in the JM, the job cannot start until a single thread had a look at the first bytes of each file.
[ Full content available at: https://github.com/apache/flink/pull/6710 ] This message was relayed via gitbox.apache.org for devnull@infra.apache.org
