Search-Based Test Data Generation for SQL Queries

Master thesis (2017)

Authors

J. Castelein Electrical Engineering, Mathematics and Computer Science

Contributors

M. Soltani (mentor)
A. van Deursen (graduation committee member)
A. Bozzon (graduation committee member)
Peter A.N. Bosman (graduation committee member)
Faculty
Electrical Engineering, Mathematics and Computer Science, Electrical Engineering, Mathematics and Computer Science
Copyright: © 2017 Jeroen Castelein
More Info
expand_more

Published Date

28-07-2017

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Abstract

Software testing is an important, well-researched field.
With the majority of modern-day applications using relational databases to manipulate their data, it is crucial that database interactions are tested as well.
This is a complex task to perform manually, and thus researchers have been attempting to tackle this problem by means of automated test data generation.
In their studies, they apply constraint-based techniques using SAT solvers to generate the test data.
However, these techniques have known limitations such as solving subqueries.

In this thesis, we present a novel search-based approach that uses a Genetic Algorithm to generate test data for SQL queries, which overcomes the limitations of previous research.
We provide an implementation of our approach, EvoSQL.
In our implementation, we instrument a real database to extract all the information necessary for the fitness function. By doing so, we support all queries using standard SQL syntax.
We evaluate our approach on 2,135 queries from 4 real-world systems, of which EvoSQL is able to cover over 96% fully.

Files

MastersThesis_JeroenCastelein.... (pdf)
(pdf | 0.414 Mb)
Unknown license