Comments on: Extending XML Document Validation with Schematron

By: Emiel Paasschens

Emiel Paasschens — Fri, 06 Jan 2012 11:00:13 +0000

Chris, you are right, performance can be an issue. Aside from the hardware, the performance is much dependent on the size en complexity of the xml file and the amount and complexity of the Schematron rules (thus size and complexity of the final xslt). For a big insurance company we 've implemented business validation of an insurance policy. The performance trick was to cache (in java) the binary result of the final xslt compilation. So for a validation only one translation has to be done, where this translation had already been compiled.

By: Chris

Chris — Thu, 05 Jan 2012 13:31:33 +0000

One of the main reasons most places don’t do plain schema validation in production is the performance impact.Â How does this impact performance?Â This looks like a really nice tool if it performs well.
Thanks
Chris

By: Emiel Paasschens

Emiel Paasschens — Thu, 20 May 2010 12:33:44 +0000

There are lots of use cases or reasons to use Schematron.Â A real-world example. For an insurance company I’ve written a validation webservice. Incoming parameters where an xml document reflecting an insurance application, name of the insurance application type and version. The application had to be validated against lots of specific insurance business rules. The webservice first validates the document against an xsd and after succesful parsing it was validated against the insurance bussines rules using Schematron. Schematron was choosen because the rules changes quite often. So these rules (in Schematron xml) were stored in a db and the webservice generated (and cached) the schematron xslt after changes in the rules (and also in case of a new version) .

By: Sumit Tambekar

Sumit Tambekar — Wed, 19 May 2010 06:47:11 +0000

Hi Emiel,
Nice article.
Schematron can be a good option if requirement says fixed business rules to be enforced on data flow.
Thanks,
Sumit Dnyanesh Tambekar