Generation of Program Analysis Tools
Frank Tip

Abstract:
In recent years, there has been a growing interest in tools that
facilitate the understanding, analysis, and debugging of programs.
This thesis is concerned with tools and techniques that support the
analysis of programs. Instead of directly implementing program
analysis tools, our aim is to generate them from formal, algebraic
specifications. This approach has the pleasant property that it is
largely language-independent. Moreover, it is shown that---to a very
large extent---the information needed to construct program analysis
tools is already implicitly present in algebraic specifications. Two
particular types of program analysis tools are studied in this thesis:
tools for source-level debugging and tools for program slicing.