DSXchange

JRodriguez wrote:Ohh! you are only parallel then this will do the trick

Convert("0123456789",'', InputCol )

Also while standardizing your data you can assign the null class to numeric token....

no... i just want to remove 77777777 if it exist in the string

JRodriguez wrote:Ohh! you are only parallel then this will do the trick

Convert("0123456789",'', InputCol )

Also while standardizing your data you can assign the null class to numeric token....

no... i just want to remove 77777777 if it exist in the string

Please post the unhandled Pattern, Pattern that you want to convert the unhandled pattern, and what you did to make this happen... I do not want to treat the unhandled pattern. Please understand that taking one unhandled pattern( specially those which contain a '+') would do unjustice to many input...

JRodriguez wrote:Oconv(InputCol, "MCA")

I guess OCONV would not help for my requirement. I looked for it. Hope CHANCE function helps

DivStand, How about posting one of those entries? As a general rule you would like to convert all unhandled pattern into pattern succesfully handled by the rule .... By the way, you can test the ressult of the Input pattern override functionality in the Rule Management GUI... one at a time is less ...

Can a particluar string/text be cleared in the transformer in DataStage

For example:
If an input field is "SAN PEDRO 7777777 ARRIBA"

and we want to remove 7777777 and produce the output as

"SAN PEDRO ARRIBA"

Can this be done? If yes, please help with the syntax.

Is this the v7.5 ruleset that you brought over to v8? Did you rename the override tables and update the PAT file where it refers to them before you brought it over? 1. No i rechecked. The original version of the rule set is 8.0. 2. In the original version the ReadOnly is in state "0" only...

The reference files don't have to be 8 chars long IIRC, but the ruleset name (ie the name of all of the main ruleset files) needs to be 8 long. As for the entries: "CENTRO CEREMONIAL" MXASENTA 850 "5 REALES" MXASENTA 850 "ANDRES CUEXCONTITLAN" "5 REALES" <-- ...

I added some 42 Input Pattern Ovverides to the Quality Stage. But even after that i do not see any change in the count and the types of Unhandled patterns. Is there any setting which needs to be done to make IP pverrides work? (i used ruleset overrides windows for the same. and provisioned all the r...

The reference files don't have to be 8 chars long IIRC, but the ruleset name (ie the name of all of the main ruleset files) needs to be 8 long. As for the entries: "CENTRO CEREMONIAL" MXASENTA 850 "5 REALES" MXASENTA 850 "ANDRES CUEXCONTITLAN" "5 REALES" <-- ...

Working with the format


"5 REALES" MXASENTA 850

instead of

"5 REALES"
MXASENTA 850


For v8.0

Working with the format


"5 REALES" MXASENTA 850

instead of

"5 REALES"
MXASENTA 850


For v8.0

JRodriguez wrote:Well ...in version 8.1 if you have an entry like yours the job will fail, even in the Rule Management GUI

Which version are you working with?

ServerVersion "8.0".

I tried keeping the new entries(only and leaving the old in 2 line format) the way you told. It worked! Thanx !!

JRodriguez wrote:Would you try to align each entry into one line?

But how does that matter? In the original rulesets, it is working perfectly this way.

Only for Standardization ...the 850 is used to indicate that uncertainty comparisons will be used so you don't need all permutation of the string By the way each entry in your lookup table is in one line, right? Like this: "SITIO EJIDO" MXASENTA 850 No it is two lines like "SITIO EJI...