Processing XML Data with DataStage

SKILL SHARING WITH MILA MASLENKOVA Today we had a terrific skill sharing session with our BI and DI consultant Mila Maslenkova about the XML Pack vs the new XML Stage in DataStage. Mila showed how to compose and read complex XML documents with the legacy XML Pack and the new XML Stage which was introduced with DataStage 8.5. We learned how the new XML Stage works and what it can do better than the legacy XML Pack, but we also learned about the limitations of the XML Stage: The new stage supports XML natively, i.e. there is no longer the need to "flatten" and recompose XML data to perform hierarchical transformations (i.e. XML to XML) like regrouping the XML document nodes. Unfortunately value calculations on text or attribute nodes (e.g. strip characters from a string, date field calculations, ...) are not directly supported in the assembly editor of the new XML stage and still require flattening of the XML data into relational data to be able to process them in a normal transformer stage. Also the XML Stage has a significantly longer startup time compared to the stages of the XML Pack. The short summary: once you understood all the operations of the new XML Stage it simplifies and makes XML processing with DataStage more consistent when compared to the old XML Pack, but on the other hand the XML Pack still has its value especially when it comes to quickly read or write small XML documents. So it is no surprise, that the XML Pack is still part of the latest DataStage release (9.1).