What Should I Code Now?

Luiz Laerte Nunes da Silva Junior (Universidade Federal Fluminense, Brazil)

Alexandre Plastino (Universidade Federal Fluminense, Brazil)

Leonardo Gresta Paulino Murta (Universidade Federal Fluminense)

Abstract: In the software development field, the amount of data related to documentation and to the source code itself is huge. Relevant knowledge can be extracted from these data, provided that the adequate tools are in place. In this context, data mining can be seen as an important tool. This paper presents a new approach for code completion based on sequential patterns mined from previous developed source code. According to what is being coded, suggestions of new code sequences are made based on the mined patterns. As a result, a plug-in for the Eclipse IDE, named Vertical Code Completion, was developed and applied over widely known Open Source systems, identifying that our approach could provide suggestions that would anticipate what a developer intends to code.

Keywords: code completion, sequential pattern mining, software maintenance

Categories: D.2.13, D.2.3, H.2.8