The objective of Phase I is to specify the scope and structure of an ADA-specific Software Testing and Analysis system, to be implemented in Phase II. The system will offer an integrated support of various software verification activities. Those fall into the categories of static and dynamic analysis techniques that share the same program data base. Static methods help identify potential software bugs by automatically analyzing the text of the program, without its execution. Prime candidates for consideration are: the detection of control and data flow anomalies, static slicing and dependency analysis. Dynamic methods are execution based and include testing, debugging and dynamic slicing. A major influence on the project will have our experience with STAD 1.0, a recently developed Pascal-based System for Testing and Debugging. However, other theories and implemented systems in public domain will also be evaluated to identify desirable features of the system. Some extensions to STAD are needed in Phase I to test the soundness of several promising new theories before their ADA implementation. At this point, two aspects of the project have to be emphasized: (1) The proposed system should have an open architecture so new theories can be easily incorporated into it and tested, and (2) The interested DoD unit should be actively involved in the project.
Benefits: (1) Reduction of software development cost due to faster fault detection and localization; (2) Increase in the reliability of software; (3) Help in automatic documentation. If successful, the results of the project can be used throughout software industry.
