Distributed Refactoring across GitHub

About this Presentation

In this session, we will identify Guava-dependent code in Github using Google Bigtable's public Github dataset, write a rule to patch the code automatically with Netflix's Rewrite, and run the rule across millions of Guava-dependency Java sources using Google Cloud Dataproc. Learn how to inoculate your code from from API changes in a proactive way!

The session will bring together three powerful technologies: Google Bigtable's daily cache of all source code in Github, Zeppelin/Spark on Google Cloud Dataproc to perform ETL at scale, and Netflix Rewrite to automatically patch code. Guava is a common source of runtime failure because of its aggressive deprecation cycle, and serves as a great demonstration of how you can use Rewrite to automatically propagate API changes through your organization's codebase (even if it isn't on Github).