A model independent source code repository
Anthony Cox, Charles L. A. Clarke, Susan Elliott Sim · 1999
Software repositories, used to support program development and maintenance, invariably require an abstract model of the source code. This requirement restricts the repository user to the analyses and queries supported by the data model of the repository. In this work, we present a software repository system based on an existing information retrieval system for structured text. Source code is treated as text, augmented with supplementary syntactic and semantic information. Both the source text and supplementary information can then be queried to retrieve elements of the code. No transformations are necessary to satisfy the requirements of a database storage model. As a result, the system is free of many of the limitations imposed by existing systems. 1 Introduction In order to store computer source code in a database, the source code must be abstracted in some manner so that it satisfies the requirements of the data model supported by the chosen database system. Databases applied to t...