A document-driven approach to database report generation

It can be argued that report generation is the most frequently performed task in database applications. Therefore the efficacy and efficiency of a database report generation mechanism has a significant impact on productivity. This paper introduces a document-driven approach as opposed to the traditional schema-driven approach to report generation resulting in more flexibility. In this approach, generating a report involves specifying it in terms of a user defined SGML-based report model. Contents of a report can be specified using a transformation language together with queries that retrieve data from different databases. A report is output as a SGML document which can be further edited as well as translated to other formats, for instance to HTML to be viewed using an Internet browser. This paper presents the approach using an example and discusses the features and usage of the transformation language which is a small but expressive language. Despite of the fact that we have only investigated the transformation against relational databases, we believe that this approach can unify report generation for different database models.