On Fri, Mar 14, 2014 at 11:06 PM, Norihiro Tanaka <[email protected]> wrote: > I changed the patch so that the delta2 shift is extracted from the trie, > because it's very excellent.
Very nice. I've begun to review these patches, and really like the improved performance. Your first version (decreasing worst-case O(M*N) to O(M+N)) gives 30x speed-up in some cases. The delta2-from-trie change brings it closer to 40x.
