I have a flow that's reading out of a database that has some text fields with UTF characters in them (windows right double quote, for example).
The data in the table is clearly the right character (UTF-8 sequence e2 80 9d) which should be escaped as \u201d but for some reason is being stored as \u001d (which is some ASCII range control character). I can't figure out where the problem is being introduced, but the JDBC driver, the database, and the destination system (MarkLogic) all support UTF. Is this a NiFi bug? Any clues on narrowing down where this is being introduced would be helpful. Thanks, Eric Ladner