Ported HttpParserBench to BenchmarkDotNet

[KrzysztofCwalina]

BenchmarkDotNet=v0.10.14.525-nightly, OS=Windows 10.0.16299.371 (1709/FallCreatorsUpdate/Redstone3)
Intel Core i7-6700K CPU 4.00GHz (Skylake), 1 CPU, 8 logical and 4 physical cores
Frequency=3914067 Hz, Resolution=255.4887 ns, Timer=TSC
.NET Core SDK=2.2.0-preview1-007460
  [Host]     : .NET Core 2.2.0-preview1-26424-02 (CoreCLR 4.6.26423.05, CoreFX 4.6.26423.04), 64bit RyuJIT
  DefaultJob : .NET Core 2.2.0-preview1-26424-02 (CoreCLR 4.6.26423.05, CoreFX 4.6.26423.04), 64bit RyuJIT

Method	Mean	Error	StdDev
RequestLine	136.0 ns	0.1786 ns	0.1583 ns
Headers	251.0 ns	1.2776 ns	1.1951 ns
FullRequest	327.8 ns	0.2701 ns	0.2526 ns

[adamsitnik]

LGTM, but will wait with merge until I get the answers

[adamsitnik]

@krwq this type is not used (and was not before the port), should we keep it or remove it?

[KrzysztofCwalina]

Yeah, I was thinking about it when I was porting the tests. I actually would like to add the struct option to these tests. But for now, we are just porting, and I will do it as a separate workitem.

[adamsitnik]

@KrzysztofCwalina do parsing methods have any side effects on Request? if not, maybe we should make it a field too?

[KrzysztofCwalina]

The code with side effects is commented out. After removing the comments the request cannot be a filed. Also, the cost of creating the request object is not avoidable, so I think it's best include it in the cost.