The SELECT clause consists of either a single identification variable that is defined in the FROM clause, or a single valued path expression that evaluates to an object reference or container managed persistence (CMP) value. You can use the DISTINCT keyword to eliminate duplicate references. For finder and select queries, the syntax of the SELECT clause is illustrated in the following example:
SELECT [ ALL | DISTINCT ]
{ single-valued-path-expression | aggregation expression | OBJECT ( identification-variable ) }
For a query that defines a finder method, the query must return an object type consistent with the home that is associated with the finder method. For example, a finder method for a department home can not return employee objects.
SELECT OBJECT(e) FROM EmpBean ej, EmpBean e WHERE ej.name = 'John' and e.salary > ej.salaryFind all departments that have one or more employees who earn less than 20000:
SELECT DISTINCT e.dept FROM EmpBean e where e.salary < 20000A select method query can have a path expression that evaluates to an arbitrary value:
SELECT e.dept.name FROM EmpBean e where e.salary < 2000
The previous query returns a collection of name values for those departments having employees earning less than 20000. A select method query can return an aggregate value:
SELECT avg(e.salary) FROM EmpBean e