Ruban shared links that lead to dead ends, but I've found the document talking about EPSM. https://arxiv.org/pdf/1012.2547v1.pdf https://arxiv.org/pdf/1012.254

Ruban shared links that lead to dead ends, but I've found the document talking about EPSM. https://arxiv.org/pdf/1012.2547v1.pdf https://arxiv.org/pdf/1012.2547v1.pdf It contains lots of benchmark results, but I don't understand how they can be useful for me, as I can't reproduce their way of benchmarking or how big the haystack really is. In fact, I don't really understand any of this. What's making matters worse, HyperScan has no binary for windows. I'd have to compile it myself, using cygwin. I can't find a precompiled binary. It irritates me that there is no easily available windows version for what's supposed to be the fastest pattern matching solution out there. But worse, it seems to be creating pattern matching files for faster matching, making it relatively useless for things like grep. On the plus side, it hinted me at an additional way of improving my solution's performance. I never considered that people might wish to save files which allow for faster pattern matching.