Model-Driven Node: Reference Implementation

This page IS part of the preceding diagram. You're running the reference implementation right now. This site was generated with 15 java classes, 15 velocity templates, 2 xsd schemas to demonstrate what an enterprise-level suite might look like with componentized node.js applications.

App A has 1152 auto-generated models (tables), collections, test data and API derived from an open source xsd schema created by the Open Travel Alliance. App B has 995 models generated from Star Standard for Automobile Retail and there's approximately 2 million test records at the moment.

I created this enterprise application in 120 hours but with a generation tool instead of by writing basic code, so now I can create a new application in a few minutes. A real e-commerce site would probably need less than 100 java classes, less than 100 velocity templates and 1 xsd schema.

The real value of the model-driven node is dynamic regeneration of all associated parts as the xsd schema evolves; the parts hold together more consistently and require less testing. In fact, in previous projects, we also generated many of our unit and integration tests using velocity templates.

"App A" Reference

    Design Docs

    Source Code

    Main Schema Models (most lookup tables filtered)

    All Schema Models

    REST API via introspection

    Kue Manager

    Kue Manager Stats

"App B" Reference

    App B