Generative And Component Based Software Engineering

This book constitutes the refereed proceedings of the Third International Conference on Generative and Component-Based Software Engineering, GCSE 2001, held in Erfurt, Germany, in September 2001. The 14 revised full papers presented together with one invited paper were carefully reviewed and selected from 43 submissions. The papers are organized in topical sections on software product lines, aspects, generic and generative approaches, and components and architectures.

This book constitutes the thoroughly refereed post-proceedings of the Second International Symposium on Generative and Component-Based Software Engineering, GCSE 2000, held in Erfurt, Germany in October 2000.The twelve revised full papers presented with two invited keynote papers were carefully reviewed and selected from 29 submissions. The book offers topical sections on aspects and patterns, models and paradigms, components and architectures, and Mixin-based composition and metaprogramming.

The size, complexity, and integration level of software systems is increasing c- stantly. Companies in all domains identify that software de?nes the competitive edge of their products. These developments require us to constantly search for new approaches to increase the productivity and quality of our software - velopment and to decrease the cost of software maintenance. Generative and component-based technologies hold considerablepromise with respect to achi- ing these goals. GCSE 2001 constituted another important step forward and provided a platform for academic and industrial researchers to exchange ideas. These proceedings represent the third conference on generative and com- nent-based software engineering. The conference originated as a special track on generative programming from the Smalltalk and Java in Industry and - ucation Conference (STJA), organized by the working group “Generative and Component-Based Software Engineering” of the “Gesellschaft fur ̈ Informatik” FG 2.1.9 “Object-Oriented Software Engineering.” However, the conference has evolved substantially since then, with its own, independent stature, invited speakers, and, most importantly, a stable and growing community. This year’s conference attracted 43 submissions from all over the world, - dicating the broad, international interest in the research ?eld. Based on careful review by the program committee, 14 papers were selected for presentation. I would like to thank the members of the program committee, all renowned - perts, for their dedication in preparing thorough reviews of the submissions.

In the past two years, the Smalltalk and Java in Industry and Education C- ference (STJA) featured a special track on generative programming, which was organized by the working group \Generative and Component-Based Software Engineering" of the \Gesellschaft fur ̈ Informatik" FG 2.1.9 \Object-Oriented Software Engineering." This track covered a wide range of related topics from domain analysis, software system family engineering, and software product - nes, to extendible compilers and active libraries. The talks and keynotes directed towards this new software engineering paradigm received much attention and - terest from the STJA audience. Hence the STJA organizers suggested enlarging this track, making it more visible and open to wider, international participation. This is how the GCSE symposium was born. The rst GCSE symposium attracted 39 submissions from all over the world. This impressive number demonstrates the international interest in generative programming and related elds. After a careful review by the program comm- tee, fteen papers were selected for presentation. We are very grateful to the members of the program committee, all of them renowned experts, for their dedication in preparing thorough reviews of the submissions. Special thanks go to Elke Pulvermuller ̈ and Andreas Speck, who proposed and organized a special conference event, the Young Researches Workshop (YRW). This workshop provided a unique opportunity for young scientists and Ph.D.

In the past two years, the Smalltalk and Java in Industry and Education C- ference (STJA) featured a special track on generative programming, which was organized by the working group \Generative and Component-Based Software Engineering" of the \Gesellschaft fur ̈ Informatik" FG 2.1.9 \Object-Oriented Software Engineering." This track covered a wide range of related topics from domain analysis, software system family engineering, and software product - nes, to extendible compilers and active libraries. The talks and keynotes directed towards this new software engineering paradigm received much attention and - terest from the STJA audience. Hence the STJA organizers suggested enlarging this track, making it more visible and open to wider, international participation. This is how the GCSE symposium was born. The rst GCSE symposium attracted 39 submissions from all over the world. This impressive number demonstrates the international interest in generative programming and related elds. After a careful review by the program comm- tee, fteen papers were selected for presentation. We are very grateful to the members of the program committee, all of them renowned experts, for their dedication in preparing thorough reviews of the submissions. Special thanks go to Elke Pulvermuller ̈ and Andreas Speck, who proposed and organized a special conference event, the Young Researches Workshop (YRW). This workshop provided a unique opportunity for young scientists and Ph.D.

Component-based software development regards software construction in terms of conventional engineering disciplines where the assembly of systems from readily-available prefabricated parts is the norm. Because both component-based systems themselves and the stakeholders in component-based development projects are different from traditional software systems, component-based testing also needs to deviate from traditional software testing approaches. Gross first describes the specific challenges related to component-based testing like the lack of internal knowledge of a component or the usage of a component in diverse contexts. He argues that only built-in contract testing, a test organization for component-based applications founded on building test artifacts directly into components, can prevent catastrophic failures like the one that caused the now famous ARIANE 5 crash in 1996. Since building testing into components has implications for component development, built-in contract testing is integrated with and made to complement a model-driven development method. Here UML models are used to derive the testing architecture for an application, the testing interfaces and the component testers. The method also provides a process and guidelines for modeling and developing these artifacts. This book is the first comprehensive treatment of the intricacies of testing component-based software systems. With its strong modeling background, it appeals to researchers and graduate students specializing in component-based software engineering. Professionals architecting and developing component-based systems will profit from the UML-based methodology and the implementation hints based on the XUnit and JUnit frameworks.