The usage of ICustomDiagnosticMessage here is non-essential - ToDiagnosticString
doesn't actually get called by ToString() in this case, due to JsonFormatter code. It was
intended to make it clearer that it *did* have a custom format... but then arguably I should
do the same for Value, Struct, Any etc.

Moving some of the code out of JsonFormatter and into Duration/Timestamp/FieldMask likewise
feels somewhat nice, somewhat nasty... basically there are JSON-specific bits of formatting, but
also domain-specific bits of computation. <sigh>

Thoughts welcome.

This fixes issue #933, effectively.

Biting off just this bit first as I don't need the changes from a previous PR for this part.

This includes all the well-known types except Any.
Some aspects are likely to require further work when the details of the JSON parsing expectations are hammered out in more detail. Some of these have "ignored" tests already.

Note that the choice *not* to use Json.NET was made for two reasons:
- Going from 0 dependencies to 1 dependency is a big hit, and there's not much benefit here
- Json.NET parses more leniently than we'd want; accommodating that would be nearly as much work as writing the tokenizer
This only really affects the JsonTokenizer, which could be replaced by Json.NET. The JsonParser code would be about the same length with Json.NET... but I wouldn't be as confident in it.

We still need the JSON representation, which relies on something like a DescriptorPool to fetch message types from based on the type URL. That will come a bit later.
(The DescriptorPool comment in this commit is just a note which will prove useful if we use DescriptorPool itself.)

While I've provided operators, I haven't yet provided the method equivalents. It's not clear to me that
they're actually a good idea, while we're really targeting C# developers who definitely *can* use the user-defined operators.

This is definitely not ready to ship - I'm "troubled" by the disconnect between a list of fields in declaration order, and a mapping of field accessors by field number/name. Discussion required, but I find that easier when we've got code to look at :)

We'll see what I've missed when CI fails...

- Move types into Google.Protobuf.Reflection
- Change codegen to reflect that in generated types

Generated code changes coming in part 2

- No parsing
- Reflection based, so not hugely efficient
- No line breaks or indentation

- Added new line at the end of SampleEnum
- Moved GeneratedMessageTest.GetSampleMessage to a new class, SampleMessages, and renamed it to CreateFullTestAllTypes.

This is mostly just making things internal instead of public, removing and reordering a bunch of code in CodedInputStream/CodedOutputStream, and generally tidying up.

- Change the default message hash code to 1 to be consistent with other code
- Change the empty list/map hash code to 0 as "empty map" is equivalent to "no map"
- Removed map fields from unittest_proto3.proto
- Created map_unittest_proto3.proto which is like map_unittest.proto but proto3-only
- Fixed factory methods in FieldCodec highlighted by using all field types :)
- Added tests for map serialization:
  - Extra fields within entries
  - Entries with value then key
  - Non-contiguous entries for the same map
  - Multiple entries for the same key

Changes to generated code coming in next commit

Generated code in next commit.

This is the start of establishing a C# namespace of "Google.ProtocolBuffers.TestProtos.Proto3" for proto3-syntax protos.
We could optionally split the directory structure as well into Proto2 and Proto3 for clarity.

This includes the NUnit test adapter which allows NUnit tests to be run under VS without any extra plugins.
Unfortunate the compatibility tests using the abstract test fixture class show up as "external" tests, and aren't well presented - but they do run.

This includes fetching the VS unit test runner package, so that tests can be run from Visual Studio's Test Explorer.

Convert both the full and lite runtimes (and json/xml serialization assemblies) to be Portable Class Libraries.
All referring projects are now .NET 4 client rather than .NET 3.5.
This commit also fixes up the ProtoBench app, which I'd neglected in previous commits. (Disentangling the two sets of changes would be time-consuming.)

This could potentially be added back in later, but its use is limited and it's a pain in terms of support in PCL environments.
One use that has been highlighted is passing objects between AppDomains; we'd recommend passing a byte array explicitly and reparsing on the other side.

This commit includes changes to the C#-specific protos, and rebuilt versions of the "stock" protos.
The stock protos have been locally updated to have a specific C# namespace, but this is expected to change soon, so hasn't been committed.
Four areas are currently not tested:
1) Serialization - we may restore this at some point, possibly optionally.
2) Services - currently nothing is generated for this; will need to see how it interacts with GRPC
3) Fields beginning with _{digit} - see https://github.com/google/protobuf/issues/308
4) Fields with names which conflict with the declaring type in nasty ways - see https://github.com/google/protobuf/issues/309

1) Remove CSharpOptions
2) A new version of DescriptorProtoFile (with manual changes from codegen - it would otherwise be Descriptor.cs)
3) Turn off CLS compliance (which we'll remove from the codebase entirely; I don't think it's actually relevant these days)
4) Add "public imports" to FileDescriptor, with code broadly copied from the Java codebase.
Lots more changes to commit before it will build and tests run, but one step at a time...