Hi Experts
I am working on change to existing file to multiple IDocs interface. We receive a file say ABC with multiple records.Currently PI is reading the file and creating Z_ABC1 and Z_ABC 2 Idocs (java/sax mapping in PI) based on below logic
Condition for ABC1 IDoc - if records S1,S2, and S3 records are together then Z_ABC1 idoc needs to be created
Condition for ABC2 IDoc - if S1,S7 and S8 segments are together then Z_ABC2 IDoc needs to be created.
Sample ABC inbound file should create 2 Z_ABC1 and 3 Z_ABC2 IDocs.
Currently the splitting logic is implemented in PI using java mapping (SAX parser) based on above ,mentioned conditions. In our design meetings we had multiple design approaches presented by team.
Approach 1 - Minimum customization need to be done at SAP and hence the IDocs needs to be split at PI and sent to SAP
Approach 2 - Business logic needs to me implemented in SAP hence this splitting should be done in SAP.
My question is:
What is the best practise to design this interface? As per my understanding Approach 1 should be apt in current scenario. Please provide your expert comments on best design approach and any other parameters need to consider here?
Sample ABC file:
Header
S1......
S2.......
S3......
S1......
S7......
S8......
S1......
S7......
S8......
S1......
S2.......
S3......
S1......
S7......
S8......
Trailer