Categories:

Conditional Expression Functions

CASE

Works like a cascading “if-then-else” statement. In the more general form, a series of conditions are evaluated in sequence. When a condition evaluates to TRUE, the evaluation stops and the associated result (after THEN) is returned. If none of the conditions evaluate to TRUE, then the result after the optional ELSE is returned, if present; otherwise NULL is returned.

In the second, “shorthand” form, the expression after CASE is compared to each of the WHEN expressions in sequence, until one matches; then the associated result (after THEN) is returned. If none of the expressions match, the result after the optional ELSE is returned, if present; otherwise NULL is returned.

Note that in the second form, a NULL CASE expression matches none of the WHEN expressions, even if one of the WHEN expressions is also NULL.

See also:

IFF

Syntax

CASE
    WHEN <condition1> THEN <result1>
  [ WHEN <condition2> THEN <result2> ]
  [ ... ]
  [ ELSE <result3> ]
END

CASE <expr>
    WHEN <value1> THEN <result1>
  [ WHEN <value2> THEN <result2> ]
  [ ... ]
  [ ELSE <result3> ]
END
Copy

Arguments

condition#

In the first form of CASE, each condition is an expression that should evaluate to a BOOLEAN value (True, False, or NULL).

expr

A general expression.

value

In the second form of CASE, each value is a potential match for expr. The value can be a literal or an expression. The value must be the same data type as the expr, or must be a data type that can be cast to the data type of the expr.

result#

In the first form of the CASE clause, if condition# is true, then the function returns the corresponding result#. If more than one condition is true, then the result associated with the first true condition is returned.

In the second form of the CASE statement, if value# matches the expr, then the corresponding result is returned. If more than one value matches the expr, then the first matching value’s result is returned.

The result should be an expression that evaluates to a single value.

In both forms of CASE, if the optional ELSE clause is present, and if no matches are found, then the function returns the result in the ELSE clause. If no ELSE clause is present, and no matches are found, then the result is NULL.

Usage Notes

  • Note that, contrary to DECODE, a NULL value in the condition does not match a NULL value elsewhere in the condition. For example WHEN <null_expr> = NULL THEN 'Return me!' does not return “Return me!”. If you want to compare to NULL values, use IS NULL rather than = NULL.

  • The condition#, expr, value, and result can all be general expressions and thus can include subqueries that include set operators, such as UNION, INTERSECT, EXCEPT, and MINUS. When using set operators, make sure that data types are compatible. For details, see the General Usage Notes in the Set Operators topic.

Collation Details

In the first form of CASE, each expression is independent, and the collation specifications in different branches are independent. For example, in the following, the collation specifications in condition1 are independent of the collation specification(s) in condition2, and those collation specifications do not need to be identical or even compatible.

CASE
    WHEN <condition1> THEN <result1>
  [ WHEN <condition2> THEN <result2> ]
Copy

In the second form of CASE, although all collation-related operations must use compatible collation specifications, the collation specifications do not need to be identical. For example, in the following statement, the collation specifications of both value1 and value2 must be compatible with the collation specification of expr, but the collation specifications of value1 and value2 do not need to be identical to each other or to the collation specification of expr.

CASE <expr>
    WHEN <value1> THEN <result1>
  [ WHEN <value2> THEN <result2> ]
  ...
Copy

The value returned from the function has the highest-precedence collation of the THEN/ELSE arguments.

Examples

This shows a typical use of CASE:

SELECT
    column1,
    CASE
        WHEN column1=1 THEN 'one'
        WHEN column1=2 THEN 'two'
        ELSE 'other'
    END AS result
FROM (values(1),(2),(3)) v;
Copy

Output:

+---------+--------+
| COLUMN1 | RESULT |
|---------+--------|
|       1 | one    |
|       2 | two    |
|       3 | other  |
+---------+--------+
Copy

This shows that if none of the values match, and there is no ELSE clause, then the value returned is NULL:

SELECT
    column1,
    CASE
        WHEN column1=1 THEN 'one'
        WHEN column1=2 THEN 'two'
    END AS result
FROM (values(1),(2),(3)) v;
Copy

Output:

+---------+--------+
| COLUMN1 | RESULT |
|---------+--------|
|       1 | one    |
|       2 | two    |
|       3 | NULL   |
+---------+--------+
Copy

This example handles NULL explicitly.

SELECT
    column1,
    CASE 
        WHEN column1 = 1 THEN 'one'
        WHEN column1 = 2 THEN 'two'
        WHEN column1 IS NULL THEN 'NULL'
        ELSE 'other'
    END AS result
FROM VALUES (1), (2), (NULL);
Copy

Output:

+---------+--------+
| COLUMN1 | RESULT |
|---------+--------|
|       1 | one    |
|       2 | two    |
|    NULL | NULL   |
+---------+--------+
Copy

The following example combines CASE with collation:

SELECT CASE COLLATE('m', 'upper')
    WHEN 'M' THEN TRUE
    ELSE FALSE
END;
+----------------------------+
| CASE COLLATE('M', 'UPPER') |
|     WHEN 'M' THEN TRUE     |
|     ELSE FALSE             |
| END                        |
|----------------------------|
| True                       |
+----------------------------+
SELECT CASE 'm'
    WHEN COLLATE('M', 'lower') THEN TRUE
    ELSE FALSE
END;
+------------------------------------------+
| CASE 'M'                                 |
|     WHEN COLLATE('M', 'LOWER') THEN TRUE |
|     ELSE FALSE                           |
| END                                      |
|------------------------------------------|
| True                                     |
+------------------------------------------+
Copy