Listing 1 - 10 of 40 | << page >> |
Sort by
|
Choose an application
Humanities and social sciences --- communications and the arts --- library science --- social sciences --- history of science --- architecture --- library science. --- history of science. --- architecture. --- Communications and the arts --- Library science. --- Social sciences --- History of science. --- Architecture.
Choose an application
Sciences and engineering --- physical sciences --- applied sciences --- computer science --- computer science. --- Physical sciences --- Applied sciences --- Computer science.
Choose an application
Sciences and engineering --- physical sciences --- applied sciences --- computer science --- computer science. --- Physical sciences --- Applied sciences --- Computer science.
Choose an application
Business management --- Science --- Higher education --- Monograph
Choose an application
Choose an application
Constraint Handling Rules (CHR), closely related to Logic Programming (LP), is a declarative programming language. Over the years, the language proved successful for implementing many kinds of problems efficiently. Mainly this, but also its simple syntax and semantics, accounts for its success and impact on the research community. To further encourage the use of CHR, we need to further improve its efficiency of execution. To this end, termination analysis of CHR can be seen as one of the main sources of input. Furthermore, in the context of program debugging, termination analysis of CHR is an important asset. Due to the many language specifics, it is often hard for programmers to point out unwanted loops in their CHR programs. It is therefore essential to have a good understanding of the termination problem in CHR. Until recently, however, there was only an informal discussion on termination of the subset of CHR that only considers simplification rules. The contributions of this thesis are therefore twofold. First, we provide for a theoretical framework for termination analysis of the full CHR language. Secondly, based on this theoretical framework, we derive an approach for automated termination analysis of CHR. This approach extends the approaches in LP to integer polynomial interpretations and can be modularised. Furthermore, the approach is practical, as we demonstrate with T*CoP, a Termination analyser for CHR on top of Prolog.
Choose an application
Choose an application
Choose an application
Choose an application
Contemporary software systems are complex compositions of diverse kinds of functionality. A key driver in reducing the cost of software development is the availability of reusable libraries. Aspect-oriented software development (AOSD) enables the localization of functionality that traditionally crosscuts the decomposition of the system. However, it remains a challenge to develop reusable aspect libraries. The reason for this is the specific nature of aspect-oriented (AO) composition. It leads to a tight and fragile coupling between aspects and base and makes it hard to manage the implicit interactions that exist between different aspects.The goal of this dissertation is to incept, describe, classify and integrate idioms and patterns for the architecture, design and implementation of reusable aspect libraries. These libraries should exhibit three reuse qualities: versatility, stability and ease-of-configuration. Versatile aspects are not coupled to any specific application, but are rather applicable to a wide range of applications. Stable aspects are both robust with respect to changes in the base code and extensible to realize new requirements. Easy configuration limits the effort and AO knowledge required to use the library in combination with a specific application.Initial research on idioms and patterns for AOSD mainly present them as independent solutions to isolated problems. They are fragmentary contributions that are not targeted towards a common goal or problem context. They lack encompassing knowledge that binds and integrates them in a coherent framework. As a result, there is little guidance in selection of the appropriate pattern, especially in the context of reusable aspect libraries.These insufficiencies form the basis of the main contributions for this dissertation. We present a structured catalog of patterns with a focus on one coherent problem context, namely, how to design reusable aspect libraries that can be easily deployed in various applications, and that are stable in the presence of evolution. We extend this catalog with a sequence of patterns that guides the development of reusable aspect libraries with easy configuration. The sequence sets out a step-wise approach and presents a first step towards a pattern language. In subsequent steps it results in an extensible and configurable core design with library-controlled mediation of the internal aspect interactions and provides multiple alternative configuration modes. Additionally, we evaluate the benefits of using the patterns and applying the pattern sequence in multiple case studies. The analyzed benefits are versatility, stability and easy configuration.In summary, the main contribution of our work is an evaluated and structured collection of patterns augmented with a step-wise approach for the architecture, design and implementation of reusable aspect libraries with easy configuration.
Listing 1 - 10 of 40 | << page >> |
Sort by
|