GitLab

Refactored emitter operator << overloads to not template them, so it's easier to overload for pointer types

Refactored the traits a bit, and added displaying the key to string and numeric key not found errors

Fixed the return value of the integral conversion functions, and also unset the dec flag so it reads other bases (just a temporary fix, since we're officially supposed to read binary too)

Modified old gcc version patch so it still uses the new Node::Read in Visual Studio. Also broke up the \uNNNN characters in the spec tests into \xNN-type strings.

Cleaned up the read template overloads (per litb's update); it seems the old version didn't compile in VS2008. Also updated the VS project files.

Refactored the operator >> and Node::Read default functions, as well as the conversion functions, to more easily read new types as keys (this uncovered an error, in example 2.11 of the spec)

Added templated casting to nodes, as well as operator == and != (for quick checks, especially to help in testing). Implemented size() on a map node to return the number of key/value pairs (as in std::map)

Converted indexing to std::size_t, and fixed the Node templated overloads to properly index any index type (determining what is an index type is a bit of a hack - it should be is_convertible<T, std::size_t> (I think), but I just explicitly wrote down a list)

Added templated Read() function that creates the output variable itself (so you don't need to have a temp variable)

Added emitting for a YAML::Node (instead of the ad-hoc std::ostream overload) so it'll actually emit valid YAML always

It's better organized now, I think - nodes only offer a single main way of getting the fundamental scalar (as a string), and now we can specialize a single template to read specific types.