The Libraries will be performing maintenance on UWSpace from July 15th-17th, 2026. UWSpace will be offline for all UW community members during this time.

Collection Disjointness Analysis in Java

Loading...
Thumbnail Image

Authors

Chu, Hang

Advisor

Journal Title

Journal ISSN

Volume Title

Publisher

University of Waterloo

Abstract

This thesis presents a collection disjointness analysis to find disjointness relations between collections in Java. We define the three types of disjointness relations between collections: must-shared, may-shared and not-may-shared. The collection- disjointness analysis is implemented following the way of a forward data-flow analysis using Soot Java bytecode analysis framework. For method calls, which are usually difficult to analyze in static analysis, our analysis provide a way of generating and reading annotations of a method to best approximate the behavior of the calling methods. Finally, this thesis presents the experimental results of the collection-disjointness analysis on several tests.

Description

LC Subject Headings

Citation