*A join in the primary read function may have caused values to be duplicate

Question

There are two logical services - Emp and Dept with getById function defined in each.
I need to design a solution; where in the user will enter the EmpId and data to update (both emp and dept table)
Structure:
Emp:
Id,Name, DeptId
Dept
Id,Name
Approach followed (Integration Layer)
declare function tns:getData($id as xs:string) as element(ns1:Details)* {
for $EMP in emp:getEmpById($id)
for $DEPT in dep:getDeptById($EMP/DEPTNO)
return
<ns1:Details>
<ns1:Emp> <ns1:Id>{fn:data($EMP/NAME)}</ns1:Id>
</ns1:Emp>
<ns1:Dept> <ns1:Name>{fn:data($DEPT/DNAME)}</ns1:Name>
</ns1:Dept>
</ns1:Details>
Now , when I create the Update Map Procedures; following Error is generated
A join in the primary read function may have caused values to be duplicated in the result. Update function is disabled.
I tired to look upon the internet; got some docs. bt couldnt figure out how to resolve this
Any help is appreciated
Rgds,
Is
Edited by: Is916 on Feb 18, 2010 11:23 PM

Answer 1

The message is telling you that the values of $EMP/NAME will be duplicated (for each $DEPT). And I'm assuming you're going to say "no they are not, since getDeptById(deptNo) returns zero or one $DEPT". But I suspect you haven't indicated that in the signature of getDeptById(deptNo). I suspect the signature looks like getDeptById( ... ) as element( dept )* - the * indicates zero or more. It should be a question mark instead of the start - indicating zero or one. You should also change the signature of getData to indicate it returns element(ns1:Details)? - since it too is going to return exactly zero or 1.
Also - please read the Best Practices announcement in this forum. parameterizing function like getEmpById(id) is not recommended. Just have getEmp() - and if you want "by id" then add a predicate .... for $emp in emp:getEmp() where $emp/ID = $id