Currently, Gepsio parses an XBRL instance and automatically loads any schemas referenced by a
<schemaRef> element. The facts in the XBRL instance are then automatically validated against the loaded schemas. However, Gepsio does not automatically load any industry-standard schemas not explicitly referenced by the
<schemaRef> element. Since XBRL instances do not typically make an explicit reference to industry-standard schemas (nor should they), facts that reference those schemas cannot be validated.
Consider, for example, the XBRL instance at http://xbrlsite.com/US-GAAP/BasicExample/2010-09-30/abc-20101231.xml. It contains one
<schemaRef> element so that a linkbase can be loaded. However, the facts in this instance reference namespaces such as
http://xbrl.us/us-gaap/2009-01-31. Since there is no schema reference for those namespaces, the facts that use those namespaces fail validation. The trick, though, is that those namespaces reference well-known industry standard namespaces that identify well-known schemas with schema source at well-known locations across the Internet.
The next version of Gepsio will add support for these well-known namespaces and will automatically load schemas referenced by the XBRL instance’s
<xbrl xmlns:...> syntax even though the schema locations for the schemas are not explicitly referenced by a
<schemaRef> element in the instance. Doing this will allow the facts in the instance using those namespaces to validate with full schema fidelity.
As of this writing, Gepsio supports the following industry-standard schemas:
- Document Information and Entity Information 2009
- US-GAAP 2009
Support for many other industry-standard schemas will be added as testing continues. The important note is that the code now has the infrastructure in place to support this notion, and support for other industry-standard schemas can be added relatively easily, and in one place, now that the infrastructure is available.