Abstract
<jats:p>This article examines the methodology of Ned Chapin's metric for assessing the complexity of software products developed using modern programming languages. The methodological aspects of the metric are presented in comparison with existing practice. The article identifies errors and incorrect approaches associated with the application of the Chapin method in software development, caused by incomplete consideration of the requirements of its methodology and the specific features of programming languages. The causes of errors arising when applying the Chapin metric are identified. Inconsistencies in data classification and the lack of consideration of program structure are highlighted. The specific features of modern programming languages are not taken into account. This paper proposes a new approach to assessing the complexity of code created in modern programming languages. The approach is based on decomposing the program into small fragments that at the same time carry a sufficiently large functional load, formalized as independent software constructs in the process description. Another feature of the proposed approach is that Chapin's method must take into account the peculiarities of modern programming languages, where the syntax of different languages provides high, but varying, line-length compactness, but can also reduce the comprehensibility of the semantics of complex expressions and dependencies. The problem of estimating the complexity of a C++ software module using Chapin's metric is considered, while adhering to the author's requirements for data classification. An example of calculating the measure's value is shown.</jats:p>