Parallel Inheritance Hierarchy: Detection from a Static View of the System

We expose a case study of a bad smell detection through metrics. In practice, bad smell detection emerges from human observations. Metrics allow to obtain an objective view of the software, so they must be used as instruments to detect bad smells. Concretely, we focus in the bad smell: Parallel Inheritance Hierarchy, using a metric subset. Although it is not a serious bad smell, however its detection is difficult in large and medium size systems. Besides, it is usually necessary to have several versions of the system to detect its presence. We define a process to manage the big amount of data extracted from a system to determine where exists this bad smell, only with an available version. The saving of time and effort in this process is showed as an advantage opposite to other solutions.