On 02/24/2013 09:46 PM, Kai Willadsen wrote: > > I think that sounds good, and is probably the easiest approach. I was > actually thinking of modifying the algorithm to require the inclusion > of the snake passing through the sync point. While that kind of > approach should be workable in a naive LCS implementation, I can't see > how we'd shoehorn that in to our current code. > > cheers, > Kai Since the current algorithm begins with common prefix/suffix matching and each sync point would actually denote a star/end of a sub-block any matching lines before/after the sync point should be recognized as a common prefix/suffix by the algorithm. So, in theory, the snake passing through the sync point should always be included.
Piotr _______________________________________________ meld-list mailing list [email protected] https://mail.gnome.org/mailman/listinfo/meld-list
