Maximizing Code Coverage in Database Applications

Abstract

A database application takes input as user-defined queries and determines the program logic to be executed based on the results returned by the queries. A change in existing application or a new application is expected pass through extensive testing to cover the entire code and check all the cases possible in execution. Testing the code coverage of traditional or CRUD-based applications is a straightforward process backed by various tools and libraries. Unlike traditional applications, checking the code coverage of database applications is a complex procedure due to its inherent structure and the inputs passed to it. Measuring the code coverage of such programs involves the participation of DBA’s to generate mock databases that can epitomize the existing data and trigger as many paths in the program as possible. Recent studies have introduced techniques to directly test the programs using the existing data. In this paper, we are comparing two methods of code coverage that will help software engineers to test their database applications without being dependent on mock databases. We aim to evaluate the performance of Baseline and Leaf Query approaches for testing such applications. Our work focuses on using decision trees as database applications and testing them by using both the methods. The efficiency of each method is assessed on several parameters by conducting experiments to understand the advantages and disadvantages of each technique.