Link Parser
"Oh happy day!" that this program should have already been made! Located at http://www.link.cs.cmu.edu/link, the program was written by Davy Temperley, Daniel Sleator and John Lafferty of Carnegie Mellon University. They have written a wonderful program with a superb API. Not only that, but there already exists a module for interface with perl: Lingua::LinkParser by Dan Brian of http://www.brians.org. All of this together makes for a fairly easy task of writing out a link parsing module in ruby.
We tried to write it from scratch, to avoid memory leaks. Release version 0.0.4 on the RAA at http://raa.ruby-lang.org/project/linkparser This pure ruby implementation now parses "He is a dog" in under 20 seconds!
Well, that's no good. Thankfully, the Abiword team fixed the memory leaks, so the regular wrapper will do fine. http://deveiate.org/projects/Ruby-LinkParser
