Sounds like we should toss the implementation asap then..
Here's a good overview of a lot of string search algorithms, with C code and pedagogic java applets showing the operation: http://www-igm.univ-mlv.fr/~lecroq/string/index.html
So someone just has to pick the best one and implement it, preferably with rsearch in mind from the start. Any takers?
(They sound mighty cool too: "Turbo Reverse Factor algorithm", "Backward Oracle Matching algorithm", or why not the "Not So Naive algorithm"?)