A formal description technique for describing transformations from one well-defined language to another is introduced. A TT-grammar contains context-free grammars for describing the syntax of both languages. The transformation between the languages is described by a relationship of productions from the grammars. The TT-grammar is supported by an automated tool. SSAGS -- a translator writing system based on attribute grammars -- has been extended to support certain classes of TT-grammars. SSAGS analyzes TT-grammars and automatically generates Ada source programs implementing the transformation specified by the TT-grammar. Experience with two different restricted forms of TT-grammars is described with respect to their practical application. The experience demonstrates the readability, ease of development, and additional verification available through the use of TT-grammars.
This paper describes the results of a study to investigate alternative techniques for using private types and packages to limit visibility of declarative items in Ada. Shortfalls of the conventional technique of applying private types, with regard to issues of recompilation, source code integrity.limiting visibility. performance. support of debugging, and interrelating abstractions, are identified.Possibly other developers of large Ada systems have encountered these issues; an analysis of over two million lines of Ada software developled for the Army, Navy, and Air Force shows that private types are rarely used. This paper describes a series of unconventional techniques for using private types and packages to layer more abstract packages on top of less abstract packages, and to develop multiple views of abstract data types applicable to different t:lasses of users. The promising techniques were applied on an actual software engineering problem, the development of a tree-builder component of an Ada-to-DIANA translator.A description of experiences, in which the techniques realized benefits for changing and debugging the software and for improving reliability, is provided.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.