Glue Language Research Articles

Python in Computational Science: Applications and Possibilities

ABSTRACT This paper focuses on the role of python in dramatic increase in productivity and high-level of code reuse in computational science. The salient features of python make it an ideal language for scientific computing exposing the shortcomings of legacy languages and prototyping platforms. Python provides a rich collection of built-in data types such as strings, lists, dictionaries; dynamic typing and dynamic binding, modules, classes, exceptions handling, automatic memory management, multiprocessing, parallel computing capabilities. Python can also be used as a glue language to wrap around existing static compiled code to obtain optimum performance. The uptrend of adopting python as a general purpose language along with its vast collection of scientific libraries are also reviewed in this paper, which ensures the long term presence of python and its growing user base in the foreseeable future. General Terms Computational Science. Keywords

Integrating three lake models into a Phytoplankton Prediction System for Lake Taihu (Taihu PPS) with Python

In the past decade, much work has been done on integrating different lake models using general frameworks to overcome model incompatibilities. However, a framework may not be flexible enough to support applications in different fields. To overcome this problem, we used Python to integrate three lake models into a Phytoplankton Prediction System for Lake Taihu (Taihu PPS). The system predicts the short-term (1–4 days) distribution of phytoplankton biomass in this large eutrophic lake in China. The object-oriented scripting language Python is used as the so-called ‘glue language’ (a programming language used for connecting software components). The distinguishing features of Python include rich extension libraries for spatial and temporal modelling, modular software architecture, free licensing and a high performance resulting in short execution time. These features facilitate efficient integration of the three models into Taihu PPS. Advanced tools (e.g. tools for statistics, 3D visualization and model calibration) could be developed in the future with the aid of the continuously updated Python libraries. Taihu PPS simulated phytoplankton biomass well and has already been applied to support decision making.

Ginga-NCL: Declarative middleware for multimedia IPTV services

This article presents the innovative features of Ginga-NCL, an open middleware specification for multimedia IPTV services. Ginga-NCL relies on the Nested Context Language, a domain-specific declarative language targeting multimedia application authoring. As a glue language, NCL relates media objects in time and space without restricting or imposing any media content type, including media objects with imperative and declarative code written using other languages. Other NCL features include support for multidevice presentations, content adaptations, presentation adaptations, and advanced code reuse. Ginga-NCL allows NCL applications to be modified on the fly by means of live editing commands. Initially defined as the standard middleware for the Brazilian terrestrial DTV system, Ginga-NCL has recently become part of ISDB standards and an ITU-T Recommendation for IPTV services.

DRuby and Rinda: implementation and application of distributed Ruby and its parallel coordination mechanism

The object-oriented scripting language Ruby is admired by many programmers for being easy to write in, and for its flexible, dynamic nature. In the last few years, the Ruby on Rails web application framework, popular for its productivity benefits, has brought about a renewed attention to Ruby for enterprise use. As the focus of Ruby has broadened from small tools and scripts, to large applications, the demands on Ruby's distributed object environment have also increased, as has the need for information about its usage, performance and examples of common practices. dRuby and Rinda were developed by the author as the distributed object environment and shared tuplespace implementation for the Ruby language, and are included as part of Ruby's standard library, dRuby extends method calls across the network while retaining the benefits of Ruby. Rinda builds on dRuby to bring the functionality of Linda, the glue language for distributed co-ordination systems, to Ruby. This article discusses the design policy and implementation points of these two systems, and demonstrates their simplicity with sample code and examples of their usage in actual applications. In addition to dRuby and Rinda's appropriateness for prototyping distributed systems, this article will also demonstrate that dRuby and Rinda are building a reputation for being suitable infrastructure components for real-world applications.

Reo: a channel-based coordination model for component composition

In this paper, we present Reo, which forms a paradigm for composition of software components based on the notion of mobile channels. Reo is a channel-based exogenous coordination model in which complex coordinators, called connectors, are compositionally built out of simpler ones. The simplest connectors in Reo are a set of channels with well-defined behaviour supplied by users. Reo can be used as a language for coordination of concurrent processes, or as a ‘glue language’ for compositional construction of connectors that orchestrate component instances in a component-based system. The emphasis in Reo is just on connectors and their composition, and not on the entities that connect to, communicate and cooperate through these connectors. Each connector in Reo imposes a specific coordination pattern on the entities (for example, components) that perform I/O operations through that connector, without the knowledge of those entities. Channel composition in Reo is a very powerful mechanism for construction of connectors. We demonstrate the expressive power of connector composition in Reo through a number of examples. We show that exogenous coordination patterns that can be expressed as (meta-level) regular expressions over I/O operations can be composed in Reo out of a small set of only five primitive channel types.

Complex Predicates and Information Spreading in LFG

Preface 1. Introduction 2. Predicate composition 3. Romance complex predicates 4. Serial verb constructions 5. Conclusions and prospects A. Glue language semantics B. Previous proposals Bibliography.