topic Best way to handle an Oracle BLOB in Talend Studio

Best way to handle an Oracle BLOB

MDouglas1687874333 — Fri, 19 Sep 2025 13:06:34 GMT

I have a setup with an oracle component that makes a call to a database and returns a BLOB datatype. The BLOB only contains textual information (I know a CLOB would work better but this is the way the database is structured) and I want to extract this as a byte array and pass this to a tFileOutputRaw.

Current skeleton/prototype:

Talend seems to want to treat it as an object, if the notification from the component and the DB TYPE being orange in the schema implies this?

This goes away when I set it to an object but it being an oracle.sql.BLOB object type, I am unsure how this should be used. I would want to get the textual value from the blob but I'm not sure how to work with this object. Any advice on this would be appreciated.

NOTE: Using a byte array works perfectly fine for this, but I am curious what, if any, benefit would be gained from using an object instead?

Re: Best way to handle an Oracle BLOB

gvarun_arthasolutions — Sat, 29 Nov 2025 20:25:15 GMT

Hello,

You’re right to notice that Talend treats Oracle BLOB columns as Java objects (oracle.sql.BLOB) unless you explicitly map them to a byte[]. That orange “DB TYPE” indicator in the schema is Talend’s way of saying “this doesn’t map cleanly to a primitive type, so we’ll expose it as an Object.”

BLOB in Oracle = Binary Large Object. It’s designed for arbitrary binary data (images, PDFs, etc.).
CLOB = Character Large Object. It’s designed for textual data.
Since your BLOB contains text, you’re essentially storing text in a binary wrapper. That’s why you need to decide how to interpret it.

Handling in Talend

Map to byte[]

In the schema, set the column type to byte[].
Talend will automatically fetch the BLOB as a byte array.
You can pass this directly to tFileOutputRaw (perfect for binary/text dumps).
If you need to interpret it as text, wrap it in new String(byteArray, "UTF-8").

Thanks