B.9 Processing component modularization and naming conventions

 

Files that implement processing associated with specific vocabulary or constraint modules should generally be organized by module, such that the implementation components correspond to the vocabulary modules that they implement and are, to the degree the implementation technology provides for it, similarly modular.

For example, the DITA Open Toolkit provides a general plugin mechanism that allows implementation components to be dynamically added to and removed from a Toolkit instance. General practice is to define Toolkit plugins that correspond to modules.

For the specific case of cascading style sheets and XSLTs transformations, two key DITA implementation technologies, the suggested filenaming conventions are:

CSS override files (recommended convention for tool providers)

XSLT override files (recommended convention for tool providers)

Previous Topic:  B.8 XML schema organization

Next Topic:  B.10 Processing interoperability considerations

Parent Topic:  Appendix B. Non-normative information

Sibling Topics:

B.1 Changes from previous versions

B.2 Migrating from DITA 1.1 to 1.2

B.3 DITA MIME type

B.4 Considerations for generalizing <foreign> elements

B.5 Element-by-element recommendations for translators

B.6 DITA module and shell implementation reference

B.7 DTD organization

B.8 XML schema organization

B.10 Processing interoperability considerations

B.11 Specialization design, customization, and the limits of specialization