Modifying Compiler Parsing to Improve the Usability of Boa: A Domain-Specific Language and Infrastructure for Ultra-Large-Scale Software Repository Data Mining

Date
2014-04-15
Authors
Srirama, Sambhav
Major Professor
Advisor
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Authors
Research Projects
Organizational Units
Computer Science
Organizational Unit
Journal Issue
Series
Department
Computer Science
Abstract

Today’s software development environment has migrated to online software repositories due to the need for a reliable collaborative space where developers can merge different parts of a complex solution. By using software repositories like GitHub or SourceForge, they can also access other publicly available projects from which they can gather valuable insight on the development process, software characteristics, etc. Unfortunately, a wealth of qualitative information is inaccessible due to the large scale nature of these repositories; only coarse-grained information can be found. The Boa Project offers fine-grained information searching functionality by allowing users to construct infrastructure queries using the Boa domain language. Previously, however, the Boa compiler technology was not able to parse incorrect code efficiently. Users were unable to get passed all the errors and exceptions due to the lack of information about what kinds of issues were interfering. By rewriting the Boa parser component using JavaCC, the compiler now has recovery and reporting functionality, allowing programmers to see specifics about what issues they are encountering and possible fixes. Additionally, the standalone component can be adapted to work within a mobile application for smartphones, which would allow users to write code and submit queries while on the move.

Comments
Description
Keywords
Citation
DOI
Source