We received an interesting question on SSDN which deserves more attentions then a simple few lines response
The original post can be found here
Here is a way to approach the problem.
First we need to add a node with an empty pattern to the custom conversion which will catch all unexpected rows.
Then we can build a schema from the converter's output which will be used to validate the document and to make sure no unexpected element are present.
At last we can build a simple XML pipeline which converts the fixed-width data to XML and performs validation against the schema.
XML Pipeline Server loads the pipeline and creates folder for each input and output blocks, it monitors the folder bound to the input and triggers the pipeline execution as soon as a new file arrives.
If you like to know more about XML Pipeline Server drop us a line at firstname.lastname@example.org
Stylus Studio Team