FACER: An API usage-based code-example recommender for opportunistic reuse

To save time, developers often search for code examples that implement their desired software features. Existing techniques typically focus on finding snippets a single given query, which means need to perform separate each functionality. In this paper, we propose FACER (F eature-driven A PI usage-based C ode E xamples R ecommender), technique avoids repeated searches through opportunistic reuse. Specifically, the selected snippet matches initial finds and suggests related represent features developer may want next. first constructs fact repository by parsing source of open-source Java projects obtain methods’ textual information, call graphs, Application Programming Interface (API) usages. It then detects unique clustering methods based similar API usages, where cluster represents feature or Finally, it frequently co-occurring across using frequent pattern mining recommends from mined patterns. evaluate FACER, run 120 Android apps GitHub. We manually validate detected method clusters with an automated evaluation determine best parameters (e.g., similarity threshold) FACER. recruit 10 professional along 39 experienced students judge FACER’s recommendation methods. Our results show that, average, recommendations are 80% precise. also survey total 20 understand reuse experiences, feedback usability usefulness The 95% our surveyed find idea useful during