Page 1 of 1
Updating reference lookups programmatically
Posted: Sat Aug 26, 2017 8:37 am
by kenstorm
I am creating a user involved process where scrubbed data needs to be approved by a SME. For data that has a low threshold, the user needs to approve or change the output then have that change inserted into the reference lookup so the next run will scrub it, return a higher threshold, and pass it through. How can I update the reference lookup within the ruleset?
Thank you for your time
Posted: Sun Aug 27, 2017 9:15 pm
by ray.wurlod
Ultimately the files (and a reference table is a file) reside in the Quality directory on the server. Therefore you should be able to modify such a file in place. However, there is no visibility of these on the client.
A better approach would be to effect the update on the client, play with the edited reference table there, and then provision it to the engine as usual.
Posted: Fri Sep 15, 2017 9:04 am
by kenstorm
Ray,
Thank you for the help!
I won't try to update those files.
I am noticing a limitation on the amount of data i can put into a reference file. At one point, i can't Provision the rule set. When i try to open the rule set from Designer, i get a "Could not load Rule Set ... [No message from repository (error 0)] error.
When i have 5450 rows in the lookup, i can provision. When i add one or more rows, i can't provision. Is this a known issue or am i overloading the rule set?
Posted: Sat Sep 16, 2017 12:03 pm
by rjdickson
No limit that I know of. In fact, ZIPSTATE.TBL has 42523 rows, and JPKOAZA00.TBL has 158438 rows.
Is the entry you are trying to add a duplicate of another one?
Posted: Mon Sep 18, 2017 7:46 am
by kenstorm
Robert,
That is what i originally thought, that there were duplicates. I searched for that one additional row that caused the issue and it wasn't a duplicate. I even tried to add a different row and got the same error. Keep in mind that i am not tokenizing, i am matching on the entire string.
Thanks - Ken
Posted: Mon Sep 25, 2017 5:13 am
by rjdickson
Something is not right. You may have to open a PMR to get IBM to help you with this.