This thesis documents the design and implementation of procedures in a database programming language. The purpose of this thesis is to integrate procedure facilities into an existing relational database system. A relation is defined over a set of attributes. Given the values of a subset of attributes as input, a selection operation looks up the relation and outputs the values of the remaining attributes. Our proposed procedure construct supports this concept: a procedure is defined over a set of parameters, and the procedure can be invoked with different subsets of input parameters. This is accomplished by allowing procedures to have a sequence of blocks within the procedure body. Each block abstracts a sequence of actions which requires a subset of parameters as input. Users can select different blocks to be activated by supplying different subsets of input parameters. While a relation can be selected with any subset of input attributes, a procedure can only be invoked wi...

Despite its immense success, the relational model of data has been underappreciated. Many wrong claims have been made to the effect that it is unable to handle complex data, to do analytical processing, or to go beyond passé, simple structured data. I have devoted most of a career in computer science to showing that relations can indeed cope with all these, without awkwardness and with minimal syntactic and conceptual extensions. Not only can relations cope; they do the job better. A further advantage of this work is integration: the same formalism that was classically used for administrative data can also be used for expert systems, for geographical information systems, for CAD-CAM, for numerical work, for data mining and for semistructured applications such as bibliographic and bioinformatic databases. Another advantage is that this integrated relational formalism is at a level of abstraction which is not only ideally suited for processing data on secondary storage but which also readily absorbs important issues in computational parallelism and in distributing data over the Internet. I review the simple ideas needed to push the relational model to its inherent full capabilities, and show the syntactic adjustments needed to avoid the limitations of conventional and commercial implementations. The discussion is prefaced by some motivating examples, without full explanations, and terminated by a consideration of some special techniques for implementing the language constructs.