CodeFill: Multi-token Code Completion by Jointly learning from Structure and Naming Sequences

Maliheh Izadi, Roberta Gismondi, Georgios Gousios

Abstract

Code completion is an essential feature of IDEs, yet current auto-completers are restricted to either grammar-based or NLP-based single token completions. Both approaches have significant drawbacks: grammar-based autocompletion is restricted in dynamically-typed language environments, whereas NLP-based autocompleters struggle to understand the semantics of the programming language and the developer's code context.

BibTeX

@inproceedings{Izadi-al:ICSE22,
  author    = {Maliheh Izadi and
               Roberta Gismondi and
               Georgios Gousios},
  title     = {{CodeFill:} Multi-token Code Completion by Jointly learning from Structure and Naming Sequences},
  booktitle = {ICSE},
  pages     = {401--412},
  publisher = {{ACM}},
  year      = {2022},
}

CodeFill: Multi-token Code Completion by Jointly learning from Structure and Naming Sequences

Abstract

Related papers