This paper describes a quantitative software testing methodology for nonstructured and structured programs. The paper first treats some of the recent work by McCabe [3] and Paige [4,5] which has developed the groundwork for a quantitative analysis on software testing. This perspective has set the stage for use of a program-graph basis as the thread for the software testing effort. A basis is a set of paths such that any other path in the graph can be expressed as a com bination of paths in the basis. A technique for generating a unique, practical basis for a program-graph is introduced. The strategy for testing programs using this basis is discussed. The final section treats the simplifying effect of structured programs on this testing approach.

In recent years, applications of graph theory to computer software have given fruitful results and attracted more and more attention. A program graph is a graph structural model of a program exhibiting the flow relation or connection among the elements (statements) in the program.

Program graphs have been used as a vehicle to focus attention on the structure of a program. A systematic methodology for partitioning a program graph (digraph) to highlight the relationships between program elements is introduced along with an attendant notation. This notation is described in purely mathematical terms in the first section, and then the programming-related implications of this approach are addressed in the second section.

With the advent of field-effect transistor (FET) technology it has become practical and economical to employ complex functions as network primitives. This paper describes a synthesis procedure for diagnosable (all single and multiple faults can be detected) FET networks. A companion procedure for generating tests to detect all faults in the resulting network is also described. This methodology does not guarantee the minimality of the network, however, it is intuitively understandable and easy to apply.