Applying Business Rules on xml files

sairamkrish · Post by **sairamkrish** » Wed Dec 13, 2006 5:27 am

Hi,

Would like to know whether it is possible to apply simple business rules like Alphanum check/If Else logic while reading from a xml file. I know that these business rules can be incorporated using a parallel transformer, but am not too sure whether i can use parallel transformer for this. Also i could see xml transformer stage in palatte. Please help me with identifying the correct stage for this.

Thanks
Krish

sairamkrish · Post by **sairamkrish** » Wed Dec 13, 2006 5:31 am

Would like to add this to my above requirement...After reading the xml file and applying the business rules i need to load the data into db2 tables on AIX.

Thanks
Krish

eostic · Post by **eostic** » Wed Dec 13, 2006 5:41 am

There are a couple of ways to answer this...first, once you've successfully read the XML document with XMLInput, then it's "just DataStage"...so you can do anything with the values of data that you would if they came from any other source........ second, you do have the option, if desired, for doing certain amounts of data validation via XML Schema. There's a performance hit, but you can check the "validation" box on the XMLInput Stage and it will perform automatic XML Schema validation (which could be doing things like datatype validation, data value against an enumerated list, checking occurs limits, etc.).... The XML Transformer is a very different stage --- it lets you invoke your own XSLT....

Ernie

roblew · Post by **roblew** » Fri Feb 16, 2007 7:38 pm

Hi, I'm sort of new to XML in general, but have requirements to build some rather simple jobs to move XML data into SQL Server, then back to XML. The transformations involved are substring, concatenating, removing/adding columns, etc.

Is there a drawback of using the XML Transform with XSLT, compared to using the parallel transformer? My instinct says that maintaining these XSLT files for business rules will be a pain in the a$$ to maintain. Therefore, I think that having any business logic/rules inside the Parallel transforms would be a better long-term solution. Plus, with the metadata repository, it would allow for better sharing, maintenance, and documentation of all business logic. Am I thinking along the correct lines?

Would there be any key benefit for using XSLT over Parallel transformations in DataStage?