Author Topic: Best Format for Multi-value I-Descriptor  (Read 3918 times)


  • Professional
  • ***
  • Posts: 45
Best Format for Multi-value I-Descriptor
« on: March 24, 2014, 02:53:59 PM »
Here is the scenario.  I am creating three I-Descriptors that will need to be multi-valued.  In a basic subroutine I begin with a single-value variable (WO.NBR).  I then read a beginning operations sequence number, create a key and look up the first operations record for that work order.  If the work center for that operation does match my filter I load my return variable with the value from a field.  I then take the Next Sequence number from that Operations Record, create a new key and read the next operations record.  Each time I match my filter and "add" another value to my return variable.  When I finish the last sequence number I leave the subroutine.

I have two questions.  First, how should I format the return variable (RESULT).  RESULT<-1>, RESULT<1,-1>, or ?  Second, one of the I-Descriptors will be as the key for the third I-Descriptor (Description).  What is the most efficient way to format this Derived (multi) values.  Another basic subroutine?


  • Rock Star
  • *****
  • Posts: 204
  • To Dream the Impossible Dream...
Re: Best Format for Multi-value I-Descriptor
« Reply #1 on: March 27, 2014, 03:19:29 PM »
I am Don Quixote so I like battling windmills. 
I see no one has come to your aid, so I will advise.
The RESULT should be based on two things 1. the number of items returned, and 2. how you plan to use the data.  For a reasonable amount of data, my gut says RESULT<1,-1>.

The other questions seem like fog,  unclear.  I would need an example.  If there is too much complexity then call a program; if not too complex just have the third 'I-descriptor' reference the second.

Onward my friend.