Choosing a configuration

Many aspects of Imvertor may be configured. Configurations take the form of XML specifications. These are based on standards and/or functional specifications. A configuration may concern aspects of the framework itself and/or particular steps.

"Configuration" thus comprises, among others:

• Choosing a metamodel, this is a collection of stereotypes and associated tagged values, as well as validation rules that are inherent to the metamodel. An important, publicly available metamodel is KKG (provided by Geonovum) and the KKG ISO profile. These are published as PDF, and the implied rules are made available as an EA profile for KKG ISO.

• Choosing a documentation format, this is a set of rules on how to compile documentation. Most notably, a "modeldoc" configuration specifies how a standard documentation approach for UML models is tweaked to align with a company's taste.

• Choosing a set of rules by which to generate XML schemas. For each schema generation approach (there are many ways to create an XML schema from one and the same UML model) a set of rules may apply.

• Choosing comparison rules. These are applied when (releases of) models are compared, which is a provided step.

• Choosing a configuration for processing UML notes. Notes may be sectioned (as in the Inspire project) and mapped onto tagged values; such mappings are processed by the framework's XMI translator.

• Choosing owner specific rules, including all visual aspects of the company (logo's, colors, fonts and the like).

And, last but not least:

• Software configuration, which is the largest part of the "configuration", and should be interpreted as dedicated plugin software modules on the Imvertor framework, that may be used to realize just about any company specific requirements. This typically concerns task-oriented XSLTs or "plugins on existing plugins" (such as validation and canonization routines), but may also include complete Java modules.