Changing order of benchmarks execution/printing

[Leszek Bartkowski]

Hi,

I am wondering if there are some options, that control the way output is
printed (and/or order in which benchmarks are executed). Currently I am
getting something like this:


Benchmark
       Mode   Samples         Mean   Mean error    Units
f.m.e.b.JmhTest.wellHelloThere
        avgt        20        0.000        0.000    ms/op
f.m.e.b.ConservativeJobExecutorBenchmarker.aggregatorBig
          ss        10    27768.926    14321.041       ms
f.m.e.b.ConservativeJobExecutorBenchmarker.aggregatorSmall
          ss        10     3016.084        3.368       ms
f.m.e.b.ConservativeJobExecutorBenchmarker.fiveRouterCircle
         ss        10   155810.011   110855.469       ms
f.m.e.b.StrictMonoJobExecutorBenchmarker.aggregatorBig
          ss        10    44050.651     5992.759       ms
f.m.e.b.StrictMonoJobExecutorBenchmarker.aggregatorSmall
          ss        10      124.168      108.465       ms
f.m.e.b.StrictMonoJobExecutorBenchmarker.fiveRouterCircle
         ss        10    10683.273     1205.494       ms

But would be nicer to have it printed like this:

Benchmark
       Mode   Samples         Mean   Mean error    Units
f.m.e.b.JmhTest.wellHelloThere
        avgt        20        0.000        0.000    ms/op
f.m.e.b.ConservativeJobExecutorBenchmarker.aggregatorBig
          ss        10    27768.926    14321.041       ms
f.m.e.b.StrictMonoJobExecutorBenchmarker.aggregatorBig
          ss        10    44050.651     5992.759       ms
f.m.e.b.ConservativeJobExecutorBenchmarker.aggregatorSmall
          ss        10     3016.084        3.368       ms
f.m.e.b.StrictMonoJobExecutorBenchmarker.aggregatorSmall
          ss        10      124.168      108.465       ms
f.m.e.b.ConservativeJobExecutorBenchmarker.fiveRouterCircle
         ss        10   155810.011   110855.469       ms
f.m.e.b.StrictMonoJobExecutorBenchmarker.fiveRouterCircle
         ss        10    10683.273     1205.494       ms

I.e. group results not by class+methodName, but just by method name, so I
can see results of executing the same methods with different implementation
right next to each other (all those methods are defined in
AbstractJobExecutorBenchmarker). Is there some way to achieve this?

Thanks,
Leszek