I just discovered the unblob project which is using hyperscan for regex matching. As hyperscan is maintained by Intel, I was wondering if some of you have an experience with it and if it could increase performance.