It takes a specification file similar to that accepted by Lex and creates a Java source file for the correspoding lexical analyzer. PR: 12476 Submitted by: Jake Burkholder <jake@checker.org>