Should a new language feature rest on evidence rather than intuition? Our ICSE 2024 paper brings the data.

April 16, 2024

Programming languages evolve, and one common way they grow is through syntactic sugar, the convenient shorthands that make code easier to read and write. Deciding which sugars to add has usually rested on a designer’s intuition, with little hard evidence about whether developers would actually benefit.

In this paper (ICSE 2024), David OBrien, Robert Dyer, Tien Nguyen, and Hridesh Rajan propose designing syntactic sugar from data. By analyzing how developers write code across a large body of real projects, the approach surfaces the patterns that occur often enough to deserve a dedicated, more readable construct, turning a language-design judgment call into an evidence-based decision.

This work is part of Analyzing Software at Scale, with Boa; for the wider story, see our overview. The full paper is available here.