Bauhaus Library

2007

2006

2005

2004

2003

Abstract: The reflexion model originally proposed by Murphy and Notkin allows one to structurally validate a descriptive or prescriptive architecture model against a source model. First, the entities in the source model are mapped onto the architectural model, then discrepancies between the architecture model and source model are computed automatically. The original reflexion model allows an analyst to specify only non-hierarchical architecture models, which is insufficient for larger systems that are decomposed into hierarchical subsystems. This paper extends the original reflexion model to hierarchical architecture models, describes a method to apply this technique, and reports on case studies conducted on two large-scale and complex applications (namely, the C compiler sdcc for 8-bit microprocessors and the GNU C compiler gcc).

2002

2001

Abstract: Understanding a system's implementation without prior knowledge is a hard task for reengineers in general. However, some degree of automatic aid is possible. In this paper, we present a technique building a mapping between the system's externally visible behavior and the relevant parts of the source code. Our technique combines dynamic and static analyses to rapidly focus on the system's parts urgently required for a goal-directed process of program understanding.

2000

Abstract: The literature is rich of fully automatic and semi-automatic techniques for component recovery and their number is still growing. The abundance of published methods calls for frameworks to unify, classify, and compare them in order to make informed decisions. This thesis introduces a classification of component recovery techniques based on a unification of 23 techniques. Focussing on structural techniques, 16 fully automatic techniques are classified into connection-, metric-, graph-, and concept-based subcategories and the commonalities and variabilities of these techniques are discussed in depth. Beyond the qualitative comparison, 12 structural techniques are evaluated quantitatively (concept-based techniques were excluded). To that end, an evaluation scheme is introduced that allows to measure recall and precision of component recovery techniques with respect to a set of reference components ascertained by software engineers. Among the evaluated techniques is our new metric-based technique named Similarity Clustering. The evaluation scheme based on a set of expected components manually compiled by 5 software engineers for four C systems with altogether 136 KLOC shows that Similarity Clustering is among the best techniques for all systems, but it also has more false positives than other techniques. The overall result of this comparison is that none of the fully automatic techniques has a sufficient detection quality.

In order to overcome this problem, a semi-automatic method is presented in this thesis in which computer and maintainer collaborate to detect components. The method is supported by a framework that integrates the existing fully automatic techniques. In this framework, the automatic techniques can be run successively and their results be validated by the user. For this purpose, all the techniques are enhanced to work incrementally. The unification of the automatic techniques makes it possible to implement incremental variants for whole classes of techniques. The results of the techniques can be combined by high-level operators modeled on intersection, union, and difference for fuzzy sets. An alternative way of integration is offered by a voting approach that summarizes the individual agreement of automatic techniques.

Despite of the new ways of combining the automatic techniques, the semi-automatic method inherits weaknesses of the integrated techniques. Future research should investigate whether these weaknesses may be overcome with additional, more precise information gained from dataflow analyses and more domain-oriented information. However, all methods will always have to cope with vagueness and subjectivity of the grouping criteria for components.

Abstract: We used the Bauhaus tool, which helps program understanding and architecture recovery, to analyze the xfig program. xfig's architecture was recovered and all of the (hypothetical) maintenance tasks described in the Developer Handbook were performed.

Abstract: In the context of the Bauhaus project, reengineering environments to support program understanding of legacy code are being developed. Bauhaus defines two formats to represent information that has been extracted from source code. One of these formats, RG, is suitable as an exchange format. This paper introduces RG, describes how it is represented as an exchange format, and discusses schema conversions in RG.

Abstract: Component recovery supports program understanding, architecture recovery, and re-use. Among the best known techniques for detection of re-usable objects (related global variables and their accessor functions) is Delta-IC. This paper re-visits the original approach and extends it in different ways. It describes a variant of Delta-IC suitable for reverse engineering that omits the slicing step of the original approach. The underlying metric of Delta-IC is extended toward types integrating ideas of the Internal Access technique such that abstract data types can also be detected. Furthermore, the connectivity metric of Delta-IC is combined with a cohesion metric based on vertex connectivity. The new metrics and the new algorithm for reverse engineering are quantitatively evaluated using the framework proposed in as a standard evaluation of clustering techniques for component recovery.

1999

1998

1997

1996