To eliminate the need to cleanse Unicode data before loading and to reduce translation errors, use UNICODE PASS THROUGH. This option allows pass through characters to be imported into and exported from Teradata. Pass through characters include:
- BMP from Unicode versions 6.1.0 to 9.0.0, which Teradata does not support, including emoji
- SMP from all Unicode versions, which Teradata does not support
- Unassigned characters
- Private use characters
UTF-8 or UTF-16 sessions with UNICODE PASS THROUGH enabled allow pass through characters to be used in requests and to be stored in or retrieved from Unicode columns. Noncharacters and characters with an invalid encoding form are changed to the REPLACEMENT CHARACTER (U+FFFD), which can pass through and be stored in Teradata with this option.
UNICODE PASS THROUGH is not supported with FastLoad, MultiLoad, Teradata Parallel Data Pump (TPump), and FastExport. Use Teradata Parallel Transporter (TPT), BTEQ, or the Teradata JDBC Driver to load or unload data containing pass through characters.
If you do not specify UNICODE PASS THROUGH for a session, the default is OFF.