To create your own collation sequence, define for each character of the alphabet the hexadecimal representation for each of the following:
- The equivalence class
- The casespecific value
- The uppercase value
It is useful to have the hexadecimal representation for the Teradata Database internal value of the character so that the preceding items can be placed in the appropriate position within the defining byte string, as shown in Example of Installing a Japanese Character Standard Collation.
Each group of values is inserted into a column in one row of DBC.CollationsV.
Site-defined collations are valid only for single-byte characters.