By default, a string key masking rule is case sensitive. This means that the rule considers an input character in uppercase and lowercase as different characters. Therefore, a successful masking task might change the case of a character but not the character. For example, a masking task might mask an input character "x" to "X."

To ensure that the masking rule considers the input character and case, choose the

Case Insensitive

option. The rule then considers an input character in uppercase and lowercase as the same character. Therefore it cannot mask an input character "x" to "X."

Source data might include the same data in different formats. For example, the data might contain a name in different ways in different tables:

John Brown
jOHN bROWN
john brown
JOHN BROWN

If you configure the rule to be case insensitive but do not convert the data to a uniform format, the task might mask the same character that occurs in different cases in the source differently in the target. In the example, instances of

john

cannot be masked to

JOHN

. However, if the data contains both

john

and

JOHN

, both instances might be masked to different values.

To ensure that a string key masking rule returns deterministic masked output for the same data regardless of format, you must convert the source data to a uniform format before a masking task applies the masking rule. Configure a preprocessing expression in the rule to convert all characters to the same format.