Automatically identifying source code relevant to a task

摘要

Mechanisms are provided to automatically identify source code relevant to a task among a set of disparate source code. The mechanisms search a plurality of working sets based on a given keyword associated with the task in order to identify a subset of working sets from a working set store that comprises the given keyword. The mechanisms identify source code files associated with the identified working sets. The mechanisms determine a score that depends on the occurrences of matched working sets for the given keyword within each identified source code file. For each directory or repository, the mechanisms determine an aggregated score for all the source code files comprised therein. The mechanisms present a listing of all identified source code files by directory and repository, where the listing illustrates a relevancy of repositories, directories, and source code files that comprise the given keyword associated with the task.