Learning Code Transformations via Neural Machine Translation

Source code evolves – inevitably – to remain useful, secure, correct, readable, and efficient. Developers perform software evolution and maintenance activities by transforming existing source code via corrective, adaptive, perfective, and preventive changes. These code changes are usually managed an...

Full description

Bibliographic Details
Main Author: Tufano, Michele
Format: Others
Language:English
Published: W&M ScholarWorks 2019
Subjects:
Online Access:https://scholarworks.wm.edu/etd/1563899006
https://scholarworks.wm.edu/cgi/viewcontent.cgi?article=6811&context=etd