eliminate Lazy<string> from Address

[Aaronontheweb]

This PR:

1. Improves ActorPath.Parse performance
2. Reduces Address memory allocation

[Aaronontheweb]

ActorPath.Parse Performance

Before

BenchmarkDotNet=v0.13.0, OS=Windows 10.0.19041.985 (2004/May2020Update/20H1)
AMD Ryzen 7 1700, 1 CPU, 16 logical and 8 physical cores
.NET SDK=5.0.203
  [Host]     : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT
  DefaultJob : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT

Method	Mean	Error	StdDev	Gen 0	Gen 1	Gen 2	Allocated
ActorPath_Parse	583.49 ns	11.631 ns	16.305 ns	0.1392	-	-	584 B
ActorPath_Concat	55.87 ns	1.142 ns	1.315 ns	0.0268	-	-	112 B
ActorPath_Equals	20.35 ns	0.336 ns	0.314 ns	-	-	-	-
ActorPath_ToString	207.66 ns	4.186 ns	3.916 ns	0.0515	-	-	216 B
ActorPath_ToSerializationFormat	225.86 ns	2.761 ns	2.583 ns	0.0515	-	-	216 B
ActorPath_ToSerializationFormatWithAddress	244.88 ns	4.940 ns	4.379 ns	0.0515	-	-	216 B

After

BenchmarkDotNet=v0.13.0, OS=Windows 10.0.19041.985 (2004/May2020Update/20H1)
AMD Ryzen 7 1700, 1 CPU, 16 logical and 8 physical cores
.NET SDK=5.0.203
  [Host]     : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT
  DefaultJob : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT

Method	Mean	Error	StdDev	Gen 0	Gen 1	Gen 2	Allocated
ActorPath_Parse	526.80 ns	8.282 ns	7.342 ns	0.1068	-	-	448 B
ActorPath_Concat	58.06 ns	1.214 ns	1.399 ns	0.0267	-	-	112 B
ActorPath_Equals	20.66 ns	0.400 ns	0.374 ns	-	-	-	-
ActorPath_ToString	210.67 ns	4.207 ns	5.616 ns	0.0515	-	-	216 B
ActorPath_ToSerializationFormat	232.91 ns	4.497 ns	4.998 ns	0.0515	-	-	216 B
ActorPath_ToSerializationFormatWithAddress	246.31 ns	4.979 ns	7.140 ns	0.0515	-	-	216 B

Address Performance

Before

BenchmarkDotNet=v0.13.0, OS=Windows 10.0.19041.985 (2004/May2020Update/20H1)
AMD Ryzen 7 1700, 1 CPU, 16 logical and 8 physical cores
.NET SDK=5.0.203
  [Host]     : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT
  DefaultJob : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT

Method	Mean	Error	StdDev	Gen 0	Gen 1	Gen 2	Allocated
Address_Parse	1,440.3613 ns	8.8500 ns	7.8453 ns	0.1354	-	-	568 B
Address_CompareTo	16.5944 ns	0.1334 ns	0.1247 ns	-	-	-	-
Address_ToString	0.6547 ns	0.0500 ns	0.0443 ns	-	-	-	-
Address_Equals	7.9953 ns	0.1255 ns	0.1174 ns	-	-	-	-
Address_GetHashCode	34.5831 ns	0.3759 ns	0.3516 ns	-	-	-	-

After

BenchmarkDotNet=v0.13.0, OS=Windows 10.0.19041.985 (2004/May2020Update/20H1)
AMD Ryzen 7 1700, 1 CPU, 16 logical and 8 physical cores
.NET SDK=5.0.203
  [Host]     : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT
  DefaultJob : .NET Core 3.1.15 (CoreCLR 4.700.21.21202, CoreFX 4.700.21.21402), X64 RyuJIT

Method	Mean	Error	StdDev	Gen 0	Gen 1	Gen 2	Allocated
Address_Parse	1,368.8126 ns	9.9937 ns	8.8591 ns	0.1030	-	-	432 B
Address_CompareTo	17.3290 ns	0.1720 ns	0.1609 ns	-	-	-	-
Address_ToString	0.6843 ns	0.0619 ns	0.0549 ns	-	-	-	-
Address_Equals	7.4430 ns	0.1869 ns	0.1836 ns	-	-	-	-
Address_GetHashCode	34.9871 ns	0.3221 ns	0.3013 ns	-	-	-	-

[Arkatufus]

This optimization might actually hurt memory footprint in use cases where there are a lot of actors and address ToString() are never called.

[Arkatufus]

Oh, I didn't notice the changes on the ToString() method, yes, this is good.