Feature: Keep request body in memory also after consuming request body. (#395 by @clue)

This means consumers can now always access the complete request body as detailed in the documentation. This allows building custom parsers and more advanced processing models without having to mess with the default parsers.