Looking for some advice on how to crack this nut.... I've tried a
couple of different approaches and reached road blocks on both.
Here is the deal.
I need to build a list of store numbers an employee can be affiliated
with. The data is stored in 5 attrs - storeNumber, storeNumber1,
storeNumber2... storeNumber4.
To make it even more complicated, each of the 5 attributes can have
multiple store numbers in pipe delimited entries. The goal is get all
the store numbers that appear in any of the 5 attributes consolidated
into the storeNumber attr (in the pipe delimited format) for passing to
the target. This is a delimited text driver, so the formatting is key.
storeNumber: 2020
storeNumber1: 2121|2533|2421
storeNumber3: 2020|2001
storeNumber4: 5151||3535|2121||

The goal is to get the storeNumber being sent to look like this:
storeNumber: 2020|2121|2533|2421|2001|5151|3535
The duplicate pipes shown above are possible in the data too.
So far, I have put all the values of all the attrs into LVs and
concatenated the LVs into one big one. This gets all the data there
into the sole storeNumber attr, but I still need to take out the

The two options, as I see it:
1. Do it all in policy with nested for-each loops to test each value
against the rest of the list, and if it appears two or more times, drop
one of the values. Convoluted and messy, as well as a performance
2. Do it in ECMA. I have a script that will de-dupe a nodeset, which
was my original idea. Then - it dawned on me, and this isn't a nodeset,
it's just a pile of unstructured data.

IS there a way to "make" a nodeset? - to trick IDM into treating this
like a nodeset? Isn't a nodeset just an instance doc, with multiple
elements with values (and some other attributes?) I suppose there is no
way to "fake" a nodeset as this is likely a dependency of the shim's
variable / class definition.
I think that the easiest way to "done" from here is to re-write the ECMA
so that it handles the data as it comes in, andI may have answered my
own question unless someone has another approach to this problem.

desertpilotjc's Profile: https://forums.netiq.com/member.php?userid=9777
View this thread: https://forums.netiq.com/showthread.php?t=53668