Database and their applications are topics of interest to both academia and industry. However, they have received little attention towards improving the knowledge about their associated faults and failures. This lack o.f knowledge is an impediment to the definition o.f adequate so.ftware testing techniques applicable in this domain and to the develop ment (�f quality so.ftware. We discuss issues arising from SQL manipulation failures and present the results of an investigation aiming at understanding the relationship be tweenfaults andfailures. The analysis (�fdata mapping in dicates that: i) there is a many-to-many mapping between faults andfailures; ii) failure dimensions are dependent on fault type, faulty command, and the database itse(t;-and iii) knowledge o.fthe manipulation faults is crucial to program ming and testing database applications.

No abstract

Active Rule databases have been used as an alternative to the partial implementation of applications in several knowledge domains. Their principle is the automatic response to events by the activation of tasks with specific functionalities, leading to the execution of active rules. Notwithstanding their widespread use, few research efforts have been concentrated on active database application testing. In this research work we investigate the use of a structural testing technique to reveal the presence of faults, aimed at improving reliability and overall quality of this kind of software.A family of adequacy criteria is proposed and analysed in the active SQL-based database realm. Specifically, an interaction model between rules is elaborated, in order to abstract interaction associations that form the basis for testing requirements. In the context of data flow based structural testing, a family of adequacy criteria is defined, called Interaction Between Rules based Criteria, that demands the coverage of interaction associations. The criteria are an extension to the all uses criterion, by the exploitation of persistent data flow relations associated to rule interaction. Both theoretical and empirical investigations were performed, showing that the criteria posses fault detecting ability with polynomial complexity.Manipulation faults and failures were studied, enumerated and used in an experiment that evaluates criteria fault detecting ability at different granularities: data flow analysis precisions. A tool called ADAPT-TOOL (Active Database APplication Testing TOOL for active rules written in SQL) was built to support the experiment. The results indicate that: i) the fault-detecting efficacy was 2/3 of the adequate set, and reaches higher values for the lower data flow analysis precision; and (ii) the coverage of interaction association at higher granularities does not improve the fault detecting ability. ivA Cyndia e a Giovanna, minha família. v AgradecimentosA Deus, pela delicadeza da vida.Ao prof. Mario Jino, pelas oportunidades, pela orientação e pelas conversas que resultaram nesta pesquisa.Ao prof. Plínio Vilela pelas contribuições a esta pesquisa.Aos meus pais, Plínio e Eliezita, pelo amor, pela educação e por sua verdadeira amizade.Aos meus irmãos, por sermos uma família.A minha esposa, Cyndia, pelo amor, dedicação e sorrisos que iluminam nossas vidas.A minha filha, Giovanna, por sua inocência e chorinho.Aos colegas da UNICAMP, em especial aos do grupo de teste de software da FEEC.