Software Metrics for Object-Oriented Software
J. Chris Coppick · TopSCHOLAR (Western Kentucky University) · 1990
Within this thesis the application of software complexity metrics in the object-oriented paradigm is examined. Several factors which may affect the complexity of software objects are identified and discussed. The specific applications of Maurice Halstead’s Software Science and Thomas McCabe’s cyclomatic-complexity metric are discussed in detail. The goals here are to identify methods for applying existing software metrics to objects and to provide a basis of analysis for future studies of the measurement and control of software complexity in the object-oriented paradigm of software development. Halstead’s length, vocabulary, volume, program levels and effort metrics are defined for objects. A limit for the McCabe cyclomatic complexity of an object is suggested. Also, tools for calculating these metrics have been developed in LISP on a Texas Instruments’ Explorer.