We don't actually need the range for layout nodes, so just store it
for token nodes. This will also make deferred node handling easier
later on, because we don't need to keep track of layout node ranges.
Previously, the passed in source file was implicitly terminated by the
first null character. The source file might, however, contain a null
character in the middle and we shouldn't stop parsing at it.
Add a shared library with a C API that provides access to the syntactic parser with callbacks for the inference of raw syntax nodes.
This is primarily intended to be used by SwiftSyntax to speed-up source code parsing for it.
