I pulled out some code from a Common Crawl project to answer an old Stack Overflow question.
The interesting file is src/main/scala/com/banshee/ccspark/RawWarc.scala. It contains code to let you open WARC files in Hadoop/Spark.
Release this as an actual library.