Write a Blog >>

The Kythe project defines a graph model for representing glosses of programs. Tools like source code browsers and documentation generators that act on this model can therefore be written only once rather than once per language. The graph model is designed to scale to very large codebases like Google’s. We support C++, Java, and Go in our open-source project (with developing support for TypeScript and Rust); internally, we have different levels of support for Python, Dart, Closure JavaScript, protocol buffers (an interface description language), and various proprietary languages. In this talk, we introduce the Kythe graph model, describe how our tools make use of that model, and present a case study on linking between languages through generated code from protocol buffers.

Tue 20 Jun

Displayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change

10:25 - 12:45
Tuesday - 10:25 - 12:45 - Sala d'ActesCurry On Talks at Sala d'Actes, Vertex Building
10:25
40m
Talk
Grammar-based language modes for text editors
Curry On Talks
11:15
40m
Talk
Scalable cross-references across languages
Curry On Talks
Luke Zarko Google, Inc
12:05
40m
Talk
There are no BFT Fans Anymore... About Secure Eventual Consistency
Curry On Talks
Ali Shoker HASLab/INESC TEC & University of Minho