The Translation Approach.
Translation is an innovative, language-independent approach to software design that applies a set of design patterns and translation rules to Executable UML models in order to produce code. This approach enables developers to build systems that possess a high degree of design regularity, automate the bulk of the implementation, and produce highly reusable software components.
At the heart of translation is the formulation of a model compiler: a packing of the system-wide rules and policies for the organization and management of data, control and processing that support the implementation of the Executable UML models.
This course shows how to select a model compiler based on the performance requirements of the application and the capabilities of the underlying implementation platform, and then how to construct an object-oriented model compiler, using C++ as the implementation language. Finally, the course shows how to translate Executable UML models into code based on the constructed model compiler.
To ensure that the design and implementation constructs in this course are thoroughly understood, they are frequently illustrated with successfully compiled C++ code. Emphasis is placed on the underlying object-oriented programming concepts, however, and not on the language itself, making the material readily accessible to those familiar with other programming languages.