Add benchmark to detect performance regressions about hyperscan-java HOT 4 CLOSED

The best thing would be to operate on the same corpora the native hyperscan benchmark tool hsbench is operating on, see https://01.org/blogs/jpviiret/2017/performance-analysis-hyperscan-hsbench

from hyperscan-java.

I can't help to notice that hyperscan-java is an order of magnitude slower in my use case. Hence I created my test cases (not the same corpora of hyperscan benchmark though), to see if there's room for improvements. Here's how to use my benchmark.

from hyperscan-java.

@toddleo can you describe your use case and provide an example on how you attempt to use hyperscan-java in a new issue? It's not clear to me from your "benchmark code" what your exact use case is.

Furthermore, it's advisable to use a benchmark framework like JMH (see here for an introduction), because JVM benchmarking is hard to get right.

from hyperscan-java.

@gliwka you are right scalatest might not be the proper tool for benchmark. I'll see what I can do with sbt-jmh (or maybe ScalaMeter) in the following days, and update description for my use case.

Nevertheless, happy new year! 🎆

from hyperscan-java.