Parse-tree annotations meet re-engineering concerns

Abstract

We characterise a computational model for processing annotated parse trees. The model is basically rewriting-based with specific provisions for dealing with annotations along the ordinary rewrite steps. Most notably, there are progression methods, which define a default for annotating the results of rewriting. There are also access methods, which can be used in the rewrite rules in order to retrieve annotations from the input and to establish annotations in the output. Our approach extends the basic rewriting paradigm with support for the separation of concerns that involve annotations. This is motivated in the context of transformations for software re-engineering where annotations can be used to implement concerns such as layout preservation and reversible preprocessing.