A common request is to convert some data between ASCII, hexadecimal, binary, etc. format. While I am generally not a fan of "catch-all" processors and would usually suggest following the NiFi precedent/'Unix tool mentality", the conversion between these encodings is a similar process, so I think a single processor with selectable input and output formats is sufficient here. In actuality, I would suggest the core functionality be captured in a library/service with pluggable implementations and then be exposed to one processor for content conversion and one for attribute conversion. If Base64 is included, this could replace the Base64EncodeContent processor (or it could at least be marked as @Deprecated).