SAW-BOT: Proposing Fixes for Static Analysis Warnings with GitHub Suggestions
Dragoş Şerban, Bart Golsteijn, Ralph Holdorp, Alexander Serebrenik
Abstract
In this experience report we present SAW-BOT, a bot proposing fixes for static analysis warnings. The bot has been evaluated with five professional software developers by means of a Wizard of Oz experiment, semi-structured interviews and the mTAM questionnaire. We have observed that developers prefer GitHub suggestions to two baseline operation modes. Our study indicates that GitHub suggestions are a viable mechanism for implementing bots proposing fixes for static analysis warnings.
Topics & Concepts
WizardComputer scienceWizard of ozStatic analysisBaseline (sea)SoftwareSoftware engineeringMechanism (biology)World Wide WebComputer securityHuman–computer interactionOperating systemProgramming languageEpistemologyOceanographyGeologyPhilosophySoftware Engineering ResearchInformation and Cyber SecuritySoftware Reliability and Analysis Research