* Increase C# default recursion limit to 100

This matches the Java and C++ defaults.

* Change compatibility tests to use execution-time default recursion limit

This way the same tests should pass against all versions, even
if the recursion limit changes. (The tests will be testing whether
different messages work, admittedly - but that's probably fine.)

Add group support for proto2

* Ported FieldMaskUtil from Java to C#

* Merged FieldMaskUtil into FieldMaskPartial

- Removed FieldMaskUtil
- Moved FieldMaskTree to root
- Updated tests

* Improved tests

- Removed internal method FieldMaskTree.GetFieldPaths
- Proof FieldMask.Paths only contains expected values

* Added FieldMaskTreeTest to Makefile

* Added FieldMaskTree to Makefile

* Compiler changes

* Generated code changes

* Library changes

* Compiler style changes

* Generated style changes

* Fix Windows build errors

* Implement changes from review

* Reintroduce proto2 check

* Compiler changes (required handling review)

* Generated code changes (required handling review)

* Library changes (required handling review

* Field presence rewrite (compiler changes)

* Field presence rewrite (generated code changes)

* Compiler comment

* IFieldAccessor.HasValue library implementation

* Remove Clear methods and default values from proto3 code (Compiler)

* Remove Clear methods and default values from proto3 code (Generated)

* Remove Clear methods and default values from proto3 code (Library)

* Fix distcheck error

* Rewrite default string values to use base64 and convert

* Library changes (IMessage2)

* Compiler changes (IMessage2)

* Generated changes (IMessage2)

* Rebased and regenerated

* Compiler changes (initialized extension)

* Generated changes (initialized extension)

* Library changes (initialized extension)

* Refactor MessageExtensions.IsRequired

* Move string default value creator and bytes default value creator back to seperate methods

* Dead code cleanup

* Fixed segmentation fault
Removed unused header method declarations

Rather than converting the proto to a ByteString again, use the
existing SerializedData property.

This is primarily for access to comments, which would be expected to be available in a protoc plugin.

The implementation has two fiddly aspects:

- We use a Lazy<T> to avoid building the map before cross-linking. An alternative would be to crosslink at the end of the constructor, and remove the calls to CrossLink elsewhere. This would be generally better IMO, but deviate from the Java code.
- The casts to IReadOnlyList<DescriptorBase> are unfortunate. They'll always work, because these lists are always ReadOnlyCollection<T> for a descriptor type... but we can't use IList<DescriptorBase> as that's not covariant, and it's annoyingly fiddly to change the field to be of type ReadOnlyCollection<T>.

This performs more testing for field descriptors built from byte
strings too, but that's mostly incidental. The chief intent is to
check that cross-linking occurs.

Related to #658 and #5007.

This potentially avoids compiler warnings.

* Edited MessageFieldGenerator to clone to backing field instead of property

* Generated C# proto code

Updated NUnit packages, removed NUnitLite added packages for dotnet and Visual Studio, changed dotnet command in buildall to dotnet test, and deleted Program.cs (because it's no longer required).

Increased `CodedInputStream.DefaultSizeLimit` to `Int32.MaxValue` to make it consistent with the Java implementation.

By default, unknown fields are preserved when parsing. To discard
them, use a parser configured to do so:

var parser = MyMessage.Parser.WithDiscardUnknownFields(true);

Add preserve UnknownFields
Compare floating point values bitwise
Add auto-generated header to C# generated files

* Add auto-generated header to C# generated files

* regenerate C# protos

 Add unknown field support for csharp

With this in place, all tests should pass.

Fixes #3725.

This is the manual code part of the Google.Protobuf library, and tests.
Some tests will fail until codegen is changed and rerun.

If messages A and B have the same oneof case, which is a message
type, and we merge B into A, those sub-messages should be merged.

Fixes #3200.

Note that I haven't regenerated all the code, as some of the protos
have been changed, breaking generation.

Note that the default behavior is still to throw an exception; you
need to opt into ignoring unknown fields.

Fixes #2838.

Previously we only rejected the tag if the tag itself was 0, i.e.
field=0, type=varint. The type doesn't matter: field 0 is always
invalid.

This removes the last of the C# conformance failures.

Note that the compatibility tests have had to cahnge as well, to
cope with internal changes. (The test project has access to
internals in the main project.)

Fixes #3209.

- Add a TryUnpack method which doesn't throw if the type is wrong
- Make GetTypeName public for easier determination of the message type

Fixes #3294.

This has one important packaging change: the netstandard version now
depends (implicitly) on netstandard1.6.1 rather than on individual
packages. This is the preferred style of dependency, and shouldn't
affect any users - see http://stackoverflow.com/questions/42946951
for details.

The tests are still NUnit, but NUnit doesn't support "dotnet test"
yet; the test project is now an executable using NUnitLite. (When
NUnit supports dotnet test, we can adapt to it.)

Note that the project will now only work in Visual Studio 2017 (and
Visual Studio Code, and from the command line with the .NET Core
1.0.0 SDK); Visual Studio 2015 does *not* support this project file
format.