.NET persistent object oriented database framework

Abstract

The majority of software systems in demand, work hand in hand with a database on the back end of the system. This comes from the fact that enterprise systems usually store and retrieve valuable data, in order to accommodate user's needs. The need to access data in an object-oriented way arose as soon as object-oriented programming became the ideal approach to produce maintainable software. Object oriented developers often face a problem of having to switch to procedural methodologies when interacting with a database, resulting in having to use low level data access in terms of rows and columns. This results in the developer having to hold back from fully focusing on the actual application being implemented, due to extra effort required to cater for a way how to retrieve data from the database, and model it in a way that fits nicely the newly developed object-oriented application. The research section in this dissertation reviews how persistence is achieved in Java, specifically in frameworks such as, Hibernate, Castor and Simple ORM, as well as persistence mechanisms using the .NET platform, specifically Microsoft Object· Spaces and also third party implementation, namely EdgeXtend for Microsoft .NET. The second part focuses on the framework implemented. It enables the user to develop applications that need to interact with database systems, without any need to be proficient in using SQL coding, as no SQL coding is required. This reduces significantly development time, due that the framework generates all SQL code on-the-fly, to retrieve, delete and save objects to and from the database. Retrieval of data is performed by using query filters, based on object's attributes. The user can also perform Join operations and link them together to other queries, using OR and AND logical operators, to build more refined queries. Using the above is further simplified because the framework is backed up by Microsoft's IDE Intellisense, and boasts a powerful advantage, that of determining errors at compile time rather than at runtime. This dissertation also provides a sample application illustrating differences in coding, between using normal relation methodologies, and using the persistence mechanism implemented.