No description available.

We present a system for the inference of various static properties from source code written in the Maple programming language. We make use of an abstract interpretation framework in the design of these properties and define languages of constraints specific to our abstract domains which capture the desired static properties of the code. Finally we discuss the automated generation and solution of these constraints, describe a tool for doing so, and present some results from applying this tool to several nontrivial test inputs. / Thesis / Master of Science (MSc)

Researchers use datasets of Question-Solution pairs to train machine learning models, such as source code generation models. A Question-Solution pair contains two parts: a programming question and its corresponding Solution Snippet. A Solution Snippet is a source code that solves a programming question. These datasets of Question-Solution pairs can be extracted from a number of different platforms. In this research, I study how Question-Solution pairs are extracted from Stack Overflow (SO). There are two limitations of datasets of Question-Solution pairs extracted from SO: (1) according to the authors of these datasets, some Question-Solution pairs contain Solution Snippets that do not solve the question correctly, and (2) these datasets do not contain the information on how Solution Snippets need to be reused, and such information would enhance the reusability of Solution Snippets. These limitations of datasets of pairs could adversely affect the quality of the code being generated by machine learning models. In this research, I conducted a qualitative study to categorize various presentations of Solution Snippets in SO’s answers as well as how Solution Snippets can be adapted for reuse. By doing so, I identified eight categories of how Solution Snippets are presented in SO’s answers and five categories of how Solution Snippets could be adapted. Based on these results, I concluded several potential reasons why it is not easy to create datasets of Question-Solution pairs. The first categorization informs that finding the correct location of the Solution Snippet is challenging when there are several code blocks within the answer to the question. Subsequently, the researcher must identify which code within that code block is the Solution Snippet. The second categorization informs that most Solution Snippets appear challenging to be adapted for reuse, and how Solution Snippets are potentially adapted is not explicitly stated in them. These insights shed light on creating better quality datasets from questions and answers posted on Stack Overflow. / Graduate

Software developers, maintainers, and testers rely on documentation to understandthe code they are working with. However, software documentation is perceivedas a waste of effort because it is usually outdated. How documentation evolvesthrough a set of releases may show whether there is any relationship between timeand quality. The results could help future developers and managers to improvethe quality of their documentation and decrease the time developers use to analyzecode. Previous studies showed that documentation used to be scarce and low inquality, thus, this research has investigated different variables to check if the qualityof the documentation changes over time. Therefore, we have created a tool thatwould extract and calculate the quality of the comments in code blocks, classes,and methods. The results have agreed with the previous studies. The quality of thedocumentation is affected to some extent through the releases, with a tendency todecrease.

Title: Projectional editor for domain-specific languages Author: Ondřej Dvořák Department: Department of Distributed and Dependable Systems Supervisor: RNDr. Michal Malohlava Abstract: Programming is a craft requiring a good tooling. One of tools selected as crucial for software development is an integrated development environment (IDE) that allows to maintain most of the general-purpose languages. Domain-specific languages grow in a popularity last years, thus it is necessary to support them by IDE as well. Not just a textual or graphical form of DSL sources is suitable for their maintenance, so does the combination of them. One of the promising approaches is represented by a novel method called a projectional editing. Its objective is to show different visualization forms of program source code, combine and manipulate with them at one place. The thought is typically realized by a projectional editor. In this thesis we design a projectional editor for domain-specific languages and provide its experimental implementation. It analyzes potential approaches to a projectional editing and designs their suitable realization in Microsoft Visual Studio. It provides a universal implementation of a projectional editor on the top of Visual Studio as well as on the top of a standalone application. Moreover, it supports...

Open source software (OSS) is typically created in a decentralized self-organizing process by a community of developers having the same or similar interests. A key factor for the success of OSS over the last two decades is the internet: Developers who rarely meet face-to-face can employ new means of communication, both for rapidly writing and deploying software. Therefore, many tools emerged that assist a collaborative software development process, including in particular tools for source code management (SCM) and version control. In the R world, SCM is not a new idea, in fact, the R Development Core Team has always been using SCM tools for the R sources; first by means of Concurrent Versions System (CVS), and then via Subversion (SVN). A central repository is hosted by ETH Zürich mainly for managing the development of the base R system. Mailing lists like R-help, R-devel and many others are currently the main communication channels in the R community. Also beyond the base system, many R contributors employ SCM tools for managing their R packages, e.g., via web-based SVN repositories like SourceForge (http://SourceForge.net/) or Google Code (http://Code.Google.com/). However, there has been no central SCM repository providing services suited to the specific needs of R package developers. Since early 2007, the R-project offers such a central platform to the R community. R-Forge (http://R-Forge.R-project.org/) provides a set of tools for source code management and various web-based features. It aims to provide a platform for collaborative development of R packages, R-related software or further projects. / Series: Research Report Series / Department of Statistics and Mathematics

Métricas de código-fonte não são novidade, mas ainda não têm sido bem exploradas no desenvolvimento de software. A maioria das ferramentas de métricas mostra valores numéricos isolados, que não são fáceis de entender porque a interpretação deles depende do contexto da implementação. Esta dissertação apresenta o software livre Kalibro Metrics, que foi desenvolvido com o objetivo de melhorar a legibilidade de métricas de código-fonte. Kalibro, ao contrário das outras ferramentas, permite que o próprio usuário crie configurações de intervalos associados a avaliações qualitativas, incluindo comentários e recomendações. Usando essas configurações, o Kalibro mostra resultados de métricas de modo amigável, ajudando: arquitetos de software a detectar falhas de projeto; gerentes de projetos a controlar a qualidade de código-fonte; usuários de software e pesquisadores a comparar características específicas do código-fonte de vários projetos. Essas configurações podem ser compartilhadas e utilizadas para intermediar discussões voltadas à evolução de critérios de avaliação baseados em métricas de código-fonte. / Source code metrics are not new, but they have not yet been fully explored in software development. Most metric tools show isolated numeric values, which are not easy to understand because their interpretation depends on the implementation context. This dissertation presents the free software Kalibro Metrics, which was developed to improve readability of source code metrics. Kalibro, unlike the current tools, allows the user himself to create configurations of thresholds associated with qualitative evaluation, including comments and recommendations. Using these configurations, Kalibro shows metric results in a friendly way, helping: software architects to spot design flaws; project managers to control source code quality; software users and researchers to compare specific source code characteristics across software projects. These configurations can be shared and used to mediate discussions focused on the evolution of assessment criteria based on source code metrics.

Atualmente, existe uma variedade de arquiteturas disponíveis não apenas para a indústria, mas também para consumidores finais. Processadores multi-core tradicionais, GPUs, aceleradores, como o Xeon Phi, ou até mesmo processadores orientados para eficiência energética, como a família ARM, apresentam características arquiteturais muito diferentes. Essa ampla gama de características representa um desafio para os desenvolvedores de aplicações. Os desenvolvedores devem lidar com diferentes conjuntos de instruções, hierarquias de memória, ou até mesmo diferentes paradigmas de programação ao programar para essas arquiteturas. Para otimizar uma aplicação, é importante ter uma compreensão profunda de como ela se comporta em diferentes arquiteturas. Os trabalhos relacionados provaram ter uma ampla variedade de soluções. A maioria deles se concentrou em melhorar apenas o desempenho da memória. Outros se concentram no balanceamento de carga, na vetorização e no mapeamento de threads e dados, mas os realizam separadamente, perdendo oportunidades de otimização. Nesta dissertação de mestrado, foram propostas várias técnicas de otimização para melhorar o desempenho de uma aplicação de exploração sísmica real fornecida pela Petrobras, uma empresa multinacional do setor de petróleo. Os experimentos mostram que loop interchange é uma técnica útil para melhorar o desempenho de diferentes níveis de memória cache, melhorando o desempenho em até 5,3 e 3,9 nas arquiteturas Intel Broadwell e Intel Knights Landing, respectivamente. Ao alterar o código para ativar a vetorização, o desempenho foi aumentado em até 1,4 e 6,5 . O balanceamento de carga melhorou o desempenho em até 1,1 no Knights Landing. Técnicas de mapeamento de threads e dados também foram avaliadas, com uma melhora de desempenho de até 1,6 e 4,4 . O ganho de desempenho do Broadwell foi de 22,7 e do Knights Landing de 56,7 em comparação com uma versão sem otimizações, mas, no final, o Broadwell foi 1,2 mais rápido que o Knights Landing. / Nowadays, there are several different architectures available not only for the industry but also for final consumers. Traditional multi-core processors, GPUs, accelerators such as the Xeon Phi, or even energy efficiency-driven processors such as the ARM family, present very different architectural characteristics. This wide range of characteristics presents a challenge for the developers of applications. Developers must deal with different instruction sets, memory hierarchies, or even different programming paradigms when programming for these architectures. To optimize an application, it is important to have a deep understanding of how it behaves on different architectures. Related work proved to have a wide variety of solutions. Most of then focused on improving only memory performance. Others focus on load balancing, vectorization, and thread and data mapping, but perform them separately, losing optimization opportunities. In this master thesis, we propose several optimization techniques to improve the performance of a real-world seismic exploration application provided by Petrobras, a multinational corporation in the petroleum industry. In our experiments, we show that loop interchange is a useful technique to improve the performance of different cache memory levels, improving the performance by up to 5.3 and 3.9 on the Intel Broadwell and Intel Knights Landing architectures, respectively. By changing the code to enable vectorization, performance was increased by up to 1.4 and 6.5 . Load Balancing improved the performance by up to 1.1 on Knights Landing. Thread and data mapping techniques were also evaluated, with a performance improvement of up to 1.6 and 4.4 . We also compared the best version of each architecture and showed that we were able to improve the performance of Broadwell by 22.7 and Knights Landing by 56.7 compared to a naive version, but, in the end, Broadwell was 1.2 faster than Knights Landing.

O crescimento da complexidade e do tamanho dos sistemas computacionais atuais suscitou um aumento na dificuldade de extração e especificação de modelos formais desses sistemas, tornando essa atividade cada vez mais dispendiosa, tanto em tempo quanto em custo. Modelos são utilizados em diversas técnicas da Engenharia de Software, com o intuito de auxiliar em processos que compreendem desde o desenvolvimento de novos softwares, até reconstrução de um sistema a partir de software legado, passando pela realização de manutenção de um software em operação. Portanto, é necessário que essas abstrações sejam confiáveis e representem fielmente o software real. Nesse sentido, a adoção de métodos formais para a construção e análise de modelos computacionais é crescente e motivada, principalmente, pela confiabilidade que os formalismos matemáticos agregam aos modelos. No entanto, a utilização de métodos formais geralmente demanda um alto investimento de recursos humanos e, consequentemente, financeiros, uma vez que a utilização de tais formalismos é condicionada ao estudo profundo de sua fundamentação matemática. Considerando-se a extensa aplicabilidade de modelos em diversas subáreas da Ciência da Computação e as vantagens advindas da utilização de métodos formais para especificar sistemas, é interessante identificar métodos e ferramentas existentes para automatizar os processos de extração e análises de modelos, em conjunto com a adoção de formalismos que possam ser utilizados por profissionais da computação que atuam na indústria de software. Dessa forma, é estimulada nesse trabalho a utilização do formalismo de Gramática de Grafos, um método formal que diferencia-se dos demais por ser intuitivo e possuir uma representação visual gráfica, o que facilita a sua compreensão e não exige um conhecimento avançado sobre o formalismo. Primeiramente, é proposta uma abordagem para a extração de modelos em Gramática de Grafos a partir de código-fonte, extraindo informações de execuções de código Java anotado. Em seguida, é apresentada uma metodologia existente para extração e análise de Gramática de Grafos a partir de Casos de Uso, juntamente com um estudo empírico realizado a fim de validar a metodologia. Por fim, são propostas possíveis verificações adicionais, a fim de extender as análises dessa metodologia. Com isso, busca-se a obtenção de modelos, descritos através do formalismo de grafos, a partir de artefatos criados nos dois pólos do processo de desenvolvimento de software, antes e depois da implementação, no sentido de viabilizar futuras comparações, no contexto de verificação de software. / The growing size and complexity of current computer systems leading to an increase in the difficulty of extraction and specification of formal models of such systems, making it increasingly expensive activity, both in time and in cost. Models are used in various techniques of software engineering in order to assist in processes that range from the development of new software, to rebuild a system from legacy software, passing for performing maintenance of software in operation. Therefore, it is necessary that these abstractions are reliable and faithfully represent the actual software. In this sense, the adoption of formal methods for the construction and analysis of models is growing and motivated mainly by the reliability that the mathematical formalism add to models. However, the use of formal methods generally demands a high investment in human resources and hence financial, since the use of such formalism is conditioned to the deep study of its mathematical foundation. Considering the extensive applicability of models in various subfields of computer science and the benefits arising from the use of formal methods for specifying systems, it is interesting to identify existing methods and tools to automate the process of extracting models, in addition to the adoption of formalism that can be used by computer professionals working in the software industry. Thus, we encourage the use of the Graph Grammar formalism, a formal method that differs from others because it is intuitive and has a graphical visual representation, making it easy to understand and does not require an advanced knowledge of the formalism. First, we propose an approach for extracting models from source code in Graph Grammar, getting information of executions of annotated Java code. Then an existing methodology for extraction and analysis of Graph Grammar from Use Cases is presented, along with an empirical study to validate the methodology. Finally, we propose possible additional checks in order to extend the analysis of this methodology. Thus, this work aims to extract models, described by the formalism of graphs, from artifacts created in the two poles of the software development process, before and after implementation, in order to allow future comparisons, in the context of software verification.

Program comprehension, i.e. to understand from its source code what a computer programdoes, is crucial for change and maintenance in software development. In this thesis, it is lookedfor innovative documentation techniques and tools that support program comprehension, butthat are also conform to agile values and principles – commonly, documentation is consideredcritical due to the agile value “working software over comprehensive documentation.”1 First,a research framework is developed that embodies detailed requisites for such techniques andtools. Apart from its internal use for examining techniques and tools subsequently obtainedfrom a literature search, this framework is intended to be likewise employed by software practitioners.Eventually, the findings of a series of survey studies conducted in an industrial softwareorganization for the primary purpose of evaluating the obtained techniques and tools are analyzed.Three innovative techniques that meet all requisites are revealed. These are regarded bypractitioners independently from the support of program comprehension as helpful for a changeimpact analysis conducted by non-developers. Therefore, a requisite deduced from the highestpriority in agile software development – customer satisfaction – is met. It says that a techniqueor tool has to directly induce a benefit for non-developer stakeholders besides the benefits forthem which are indirectly induced by the support of program comprehension, e.g. a potentiallyimproved source code quality. Further, the technique most beneficial for developers as well asfor non-developers among the three techniques is identified, which bases on design rationales– textual information related to the source code that states the reasons why a part of the programhas been implemented in a certain way. Secondarily, the studies revealed that the researchframework is difficult to understand for practitioners due to its unstructured form. / Program: Magisterutbildning i informatik