• Reflection code has been rewritten for improved performance and simplicity. Successful extractions perform a little over 2× faster than previously. Failed extractions, at worst, perform 2× faster, and at best, perform in under 100ns (best case assumes the case path has made a successful extraction at some point of the application's lifetime). Note: some of these changes are breaking (see #32 for more details and upgrade paths), but should not affect most users of the library.
• CasePath now conforms to CustomStringConvertible.