P2R: Implementation of Processing in Racket

Processing is a programming language and development environment created to teach programming in a visual context. In spite of its enormous success, Processing remains a niche language with limited applicability outside the visual realm. Moreover, architects that have learnt Processing are unable to use the language with traditional Computer-Aided Design (CAD) and Building Information Modelling (BIM) applications, as none support Processing. In the last few years, the Rosetta project has implemented languages and APIs which enable programmers to work with multiple CAD applications. Rosetta is implemented on top of Racket and allows programs written in JavaScript, AutoLISP, Racket, and Python, to generate designs in different CAD applications, such as AutoCAD, Rhinoceros 3D, or Sketchup. Unfortunately, Rosetta does not support Processing and, thus, is not available to the large Processing community. In this paper, we present an implementation of Processing for the Racket platform. Our implementation allows Processing to use Rosetta’s APIs and, as a result, architects and designers can use Processing with their favourite CAD application. Our implementation involves compiling Processing code into semantically equivalent Racket source code, using a compiler pipeline composed of parsing, code analysis, and code generation phases. Processing’s runtime is implemented purely in Racket, allowing for greater interoperability with Racket code.