Mapping Issue for IDoc to JDBC interface

Question

Hi All,
  I am having problem in implementing logic in IDoc to JDBC interface where I have to filter out E1WBB07-KSCHL = VKP0.
  Source IDoc structure is like ->
E1WBB01(occ 0 -1000)
  |-> E1WBB03 (occ 0-100)
            |-> E1WBB07(occ 0-1000)
                     |-> KSCHL
                          DATAB
                          DATBI
Now, For each KSCHL = VKA0 there should be a duplicate VKP0 record. From these 2 records only the VKA0 should get processed and VKP0 ignored.
Duplicates for VKP0 and VKA0 can be identified by identical DATAB and DATBI.
Suppose, in one  E1WBB03 segment,there are 4 E1WBB07 segment having following values.
1: KSCHL=VKP0, DATAB=20102011, DATBI=25102011
2: KSCHL=VKP0, DATAB=26102011, DATBI=30102011
3: KSCHL=VKA0, DATAB=26102011, DATBI=30102011
4: KSCHL=VKP0, DATAB=01112011, DATBI=31129999
2 & 3 are duplicates. From these, 2 should get dropped.
As a result only 1, 3 and 4 should get  processed.
How can I proceed with this..?...I have tried some work around but not able to do it successfully. Is a UDF required to compare DATAB and DATBI. If yes how it can be written.?
Thnx in advance,
Praveen.

Answer 1

chk below mapping:
change the context of DATAB, DATAB1 and KSCHL to E1WBB03 (right click-> context) in all the mappings shown below
1)
DATAB
------------concat-----sort----splibyvalue(value change)-----collapse context---TargetNode
DATBI
2)
DATAB
------------concat-----sortbykey \
DATBI                    /        \
KSCHL------------/                 \
----------------------------------------FormatByExample----sort-----UDF1----Target KSCHL
DATAB                                    /
----concat---sort--splibyvalue(value change)-
DATBI
3)
DATAB
---concat ( ; )-----sort-splibyvalue(value change)---collapse context--splitbyvalue (each value)--UDF2---TargetDATAB
DATBI
4)
DATAB
------------concat ( ; )-----sort----splibyvalue(value change)-----collapse context--splitbyvalue (each value)--UDF3---TargetDATABI
DATBI
UDF1: execution type : all values of a context...input: var1
int a=var1.length;
int count=0;
if(a>=2)
for(int i=0;i<a;i++)
if(var1<i>.equals("VKA0"))
count= count+1;
else
result.addValue(var1[0]);
if(count>1)
for(int i=0;i<count;i++)
result.addValue("VKA0");
UDF2:execution type: single value...input: var1
String [] temp= var1.split(";");
return temp[0];
UDF3: execution type: single value...input: var1
String [] temp= var1.split(";");
return temp[1];