Rules for Writing UDFs that Compress/Decompress TIME, TIMESTAMP, Period, ARRAY, and non-LOB UDT Data - Advanced SQL Engine - Teradata Database

Follow these guidelines to write scalar UDFs that compress/decompress column data which have the following data types: TIME, TIMESTAMP, Period, ARRAY, or supported distinct non-LOB UDT.

• The compress UDF should have the following signature:
  • The UDF must declare a single input parameter which has a data type of TIME, TIMESTAMP, Period, ARRAY or supported distinct non-LOB UDT.
  • The parameter data type and column data type should be the same.
  • The return type should be VARBYTE(64000).
• The decompress UDF should have the following signature:
  • The UDF must declare a single VARBYTE(64000) input parameter.
  • The output of the decompress UDF should be identical to the column type.