DSXchange

Hi All,
I am trying to create a data Qualitystage standardize stage in DS 8.5 to standardize company names as follows:

Reference table sample from , RgTest.TBL :
"CORPORATION" CORP
"PARTNERSHIP" PART
"PARTNERS" PART
"COMPANY" CO
"FUND" FD
"BANK" BK

These rules should standardize company names of any length and apply to each word in the name - eg:
1 B 3 D E
A FUND 1 COMPANY
A PARTNERS C
PARTNERSHIP B
22 PARTNERSHIP
3 B PARTNERS D BANK
A B C D E A B C D E A B C D E PARTNERS
.. etc - you get the idea.

; (SEPLIST and STRIPLIST are standard, both include the space ' ')

?
CONVERT [1] @RgTest.TBL TEMP
COPY_S [1] {converteddata}

It works for single word name, but the "?" class doesn't seem to apply the rules to more than one word for some reason?

Do I have to have multiple "&" classes for each possible length of company name, and then apply the same CONVERT [1] to each token [2],[3],[4] ... and then CONCAT them back together?
Or is the way just to use "?" and apply to each and any word?

Please let me know how to achieve the above requirement?

Thanks,
Ram.S

Why not just add Classification overrides to your xxNAME rule set? Class these words as "O" (organizational name suffix word).

As far as I understand, using Classification overrides would not allow us to have a rule table stored in an external file, called at runtime?

Our reasons for the table approach are so we can update the rules file with new rules without recompiling the long list of jobs which use it. Does that make sense?

(I'm a colleague of the OP, Ram.S)

It does not. The only time you need to compile a job with a QualityStage rule set is if change the output (the dictionary). Please try as Ray suggested, 'provision all' the rule set, and run without compilation. Overrides are provisioned with the rule set.

Good luck!