Refactored the request charset decoding process. Added RequestID.decode_query to do the variable splitting and charset decoding in one step (scan_for_query is now a compatibility wrapper for it). Make use of the old RequestID.input_charset variable - now it actually contains the charset that has been figured out. Improved docstrings for several RequestID variables. Recognize that ISO-8859-1 is the identity mapping in join_charset. Minor cleanups in set_output_charset and charset_function. Rev: server/base_server/prototypes.pike:1.207