Interesting code. I also use a hashtable for a step of my program, but only one of my strings is hashed. That algorithm runs quickly, but isn't as detailed, it seems - I ran it on two versions of code that I had refactored, and while the changes were small (renamings), it picked up entire lines which amounted to quite a bit more than was changed.
Your zip was missing some of the files that should have been found in your .util package, but I got around that.
I'd like to release my code under a non-commercial license, but I'm unsure as to how novel the algorithm is (if at all). If it turns out to be, and is useful, I suppose I'd still have the chance to protect it against commercial use later on.
Any suggestions for a non-commercial license?
Oh, and I've updated that page ( http://katcharov.dyndns.org:8080/
) to include an example of LCS output that may be compared against the output of my algorithm.
[ October 09, 2005: Message edited by: Maxim Katcharov ]