Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.798656
Title: Supporting modern code review
Author: Han, DongGyun
ISNI:       0000 0004 8508 1208
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Modern code review is a lightweight and asynchronous process of auditing code changes that is done by a reviewer other than the author of the changes. Code review is widely used in both open source and industrial projects because of its diverse benefits, which include defect identification, code improvement, and knowledge transfer. This thesis presents three research results on code review. First, we conduct a large-scale developer survey. The objective of the survey is to understand how developers conduct code review and what difficulties they face in the process. We also reproduce the survey questions from the previous studies to broaden the base of empirical knowledge in the code review research community. Second, we investigate in depth the coding conventions applied during code review. These coding conventions guide developers to write source code in a consistent format. We determine how many coding convention violations are introduced, removed, and addressed, based on comments left by reviewers. The results show that developers put a great deal of effort into checking for convention violations, although various convention checking tools are available. Third, we propose a technique that automatically recommends related code review requests. When a new patch is submitted for code review, our technique recommends previous code review requests that contain a patch similar to the new one. Developers can locate meaningful information and development context from our recommendations. With two empirical studies and an automation technique for recommending related code reviews, this thesis broadens the empirical knowledge base for code review practitioners and provides a useful approach that supports developers in streamlining their review efforts.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.798656  DOI: Not available
Share: