Issue with Python oracledb - 3.3.0 package

[hprakasam]

Currently we are using cx-Oracle - 8.3.0 Oracle Python package to connect Oracle database with our internal application and it is deprecated. We are trying to upgrade the package to oracledb - 3.3.0. After upgrading our jobs are failing with below error while testing.
"Exception in executing Job: 'utf-8' codec can't decode byte 0xa0 in position 4: invalid start byte"
"Exception in executing Job: Error in writing to local file: 'utf-8' codec can't decode byte 0x96 in position 11: invalid start byte".

The upgraded version (oracledb - 3.3.0) is supporting UTF-8 encoding not the UTF-16, but our application is required UTF-16 encoding. Please provide suggestion on this.

[anthony-tuininga]

Can you share the relevant code that is failing? When it comes to encoding, if you are using Python strings, the conversion to/from UTF-8 for transfer to/from the database is automatic. If, however, you used cx_Oracle and specified an encoding that matched the database, BUT you actually used a completely different encoding (like UTF-16) when sending the data to the database, then this conversion will fail. There are ways to bypass the automatic encode/decode for troubleshooting purposes and to work around corrupted databases. This is documented here: https://python-oracledb.readthedocs.io/en/latest/user_guide/sql_execution.html#fetching-raw-data and in the section immediately following this one.