Originally posted by: SystemAdmin
Once again there are several ways to do this. I would agree with the earlier post that the easiest way to this may be a combination of CHOICE group and component rules.
I have a tree that has these groups defined and a map that has seperate functional maps to split them out based on the rules (the four types). It may be easier for me to send you the tree/map than to detail it here (let me know if you want them).
THe general concept to take from this example is that when you set up your choices you will want to define your hardest to satisfy group first.
I have a choice group (Pickem) containing groups:
1. xxyyaabb
2. xxaabb
3. xxyy
4. xx
The groups are all <NL> terminated and / prefix delimited. Groups xxyyaabb and xxaabb contain a mandatory repeating subgroup aabb which is / infix delimited.
In order to distinguish between groups xxyyaabb and xxaabb I have them both defined as containing element xx and yy and mandatory repeating group aabb, but in xxaabb the yy component is optional and a component rule specifies ABSENT($). In group xxyyaabb the yy component is mandatory with a PRESENT($) rule on it.
This should give you a glimpse of one way to define this data. As I said, if you would like the tree and map I was playing with, let me know.
Cheers!
#IBMSterlingTransformationExtender#DataExchange#IBM-Websphere-Transformation-Extender