Concept location in formal specifications

Abstract

When kept up-to-date, formal specifications can act as valid artifacts for maintenance tasks. However, their linguistic density and size impede comprehension, reuse, and change activities. Techniques such as specification slicing and chunking help in reducing the number of relevant lines of text to be considered, but they expect the point of change to be known a priori. This contribution presents a process model for concept location within formal Z-specifications. It also considers those situations when the location is not even roughly known. The identification is comparable to the identification of regions with high cohesion. The approach is based on the idea of first transforming the specification to an augmented graph and, secondly, on the generation of spacial clusters. A. BOLLINFormal specifications are complex 'entities'. Even simple textbook specifications (of about 20-200 lines of specification text) typically contain dozens of dependencies between the related specification elements. Within the scope of this work, these elements are called prime objects and they are, like program statements, the basic construction elements [6].In order to guarantee the benefits of formal specifications (during evolutionary steps taking place throughout the life cycle), it gets necessary to evolve and also to change the specification. This change requires to first comprehend the specification. This can be demanding, and it gets really difficult when the maintenance personnel is not the author of the specification.There are several ways of dealing with the compactness of specifications. When the location of the specification text to be changed is already known, smaller parts of the specification can be detached by slicing or chunking [7][8][9][10]. Specification slices and chunks are well-defined types of abstraction, and for Z -specifications [11] they do reduce the size of the text by 20% up to 60% [12].However, typically, a change request might not contain a reference to the lines of specification text to be altered. Requests are often formulated in terms of changing/adding/deleting features or concepts [13]. For that reason concept and feature location play an increasing role in software maintenance, and it will also play an important role during specification maintenance activities.This contribution introduces an approach facilitating the comprehension process of complex formal specifications. Research dealing with specification slices and chunks so far assumes that either the point of change is known a priori, or that this point is to be found in an intuitive way (by linearly scanning through the specification text). The main contribution of this work consists of two closely related parts: firstly, a process model for specification comprehension, and secondly, fit into it, an approach for specification concept identification aiming at exactly those cases when not even a surmise can be made about the concepts' locations.The contribution is structured as follows: Section 2 summarizes the idea behind concept location and...

Formal specification comprehension: the art of reading and writing z

“…[6, p.65] for Z), but for understanding the huge amount of text (and a lot of dependencies between different terms in the specification text) the help of tools [2] is needed. Moving away from the structure level, we do not have so much support from specification languages.…”

Coupling-based transformations of Z specifications into UML diagrams

“…Their idea is to look for predicates that are part of pre-and postconditions and to introduce "control" dependencies between them. Their idea has been refined and extended by Bollin [1] which also led to the development of an environment called V i Z that now supports reverse engineering of formal Z specifications by slicing, chunking and clustering techniques [3].…”

“…Figure 1 presents the result for the transformation of the "Access Control Specification" as done by the V i Z environment. 3 V i Z generates an X M L file that can be imported by UML modeling tools like V isual Paradigm. 4 According to the mapping rules the system class contains all operations and the initialization schema as a constructor.…”

“…A description of the mapping rules (also covering activity diagrams) including Java-like pseudo-code can be found in[21] 3. The specification is also used by other authors to demonstrate their mapping strategies and has thus been selected for this contribution 4.…”

Predictive Software Measures based on Z Specifications - A Case Study

“…Recently, Bollin showed that coupling and cohesion based measures can reasonably be mapped to formal Z specifications [3]. The basis for the calculation of all the measures is a graph that contains vertices (called primes) for all predicates and declarations of the specification, and arcs representing (reconstructed) control and data dependencies [2]. With such a graph as a basis, the following measures (defined for schemas ψ that are part of a specifications Ψ) are used in the remainder of this work: Semantics-based measures can be calculated by generating slices.…”