If I recall correctly, you get base64-encoded output from the ES connector
when you configure it to use the mapper attachment. You obviously will
want the mapper attachment installed if you are going to run the connector
in this mode.
If you are using the Tika extractor, though, I think you can just uncheck
the "use mapper attachment" checkbox in the ES connection and you will get
un-encoded utf-8 text.
On Mon, Feb 12, 2018 at 3:06 AM, Nikita Ahuja <[EMAIL PROTECTED]> wrote: