I know the title is pretty vague, however, I'll try to explain my process.
I retrieve 11 details from the database. Within each detail, there's an SSN. I need to run the validation in this order while writing an error and excluding the erroneous record from further processing:
1. SSN is null or < 9 characters
2. SSN matches a ssn in the FULL SSN exclusion list
3. SSN area matches an area in the AREA SSN exclusion list
4. SSN group matches a group in the GROUP SSN exclusion list
5. SSN serial matches a serial in the SERIAL SSN exclusion list
So for example, it item 1 is true, then write an error and ignore those records from 2-5 processing.
If item 2 is true, then write an error and ignore these records from 3-5 processing
SO at the start of the processing, I wipe away all ssn-related errors so that this is a fresh run. I identify the first step, which is an invalid SSN by way of null or < 9 characters. I flag those with an invalidSsnFlag. Now to write an error to the database for only these records with that flag, I need to filter by this flag, or remove all non-invalid SSN records, so that when I do my ADCWrite, it only writes those two records.
The only way I can see how this can progress successfully is to remove the non-valid ssns, write this to the database, do another READ call to pull in the details again, excluding those that already have an SSN-type error. I would have to do this for all 5 categories which seems to me a big overhead. Is there a different approach to do this while still using ADC? (By the way, I'm only using ADC)
Perhaps you can add a simplified Corticon Rule project Archive file to this thread of your question, explaining your question?