On Synthesizing Code from Free-Form Queries

We present a new code assistance tool for integrated development environments. Our system accepts free-form queries allowing a mixture of English and Java as an input, and produces Java code fragments that take the query into account and respect syntax, types, and scoping rules of Java as well as statistical usage patterns. The returned results need not have the structure of any previously seen code fragment. As part of our system we have constructed a probabilistic context free grammar for Java constructs and library invocations, as well as an algorithm that uses a customized natural language processing tool chain to extract information from free-form text queries. We present the results on a number of examples showing that our technique can tolerate much of the flexibility present in natural language, and can also be used to repair incorrect Java expressions that contain useful information about the developer's intent.