GitLab

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.

Re-added the throw() specification to ~Exception(), and also to ~TypedKeyNotFound(); I suppose this'll fix the gcc compiler error.

Removed throw() specifier in Exception (I don't remember putting it in). This may solve a gcc error (I haven't tested it yet) or it may break it further.

Added a templated derived exception to KeyNotFound so that you can figure out *which* key wasn't found.

(Thanks to Vadim Zeitlin)

Added Read() functions for Node that return true/false, so we can easily check if a read is successful without throwing.
But we still have operator >> that throws on failure.

Added CMake scripts for other platforms\nFixed some bugs that gcc complained about\nFixed CR/LF vs LF bug

No commit message

Changed the public interface of Scanner to resemble an STL container.

Switched the Iterator implementation to a dedicated helper class  (to hide the specific implementation, since it's pretty messy and may change).

We were getting memory leaks (as told by the CRT detectors, which I also added), and there's really no reason (as long as we're careful) to use pointers there.