I am quite happy with UTF support in phobos, but support for national codepages is very limited in phobos. Also it is not conform with ranges. So I decide share my project for supporting national charsets: https://bitbucket.org/sibnick/national-encoding.git

Sample code:

        import national.charsets, std.array;
	dstring s = "123Я";
	auto cp1251 = Windows1251.encode(s).array;
	auto utf32 = Windows1251.decode(cp1251).array;

	ubyte[] cp866 = CHARSETS["cp866"].encode(s).array;
	utf32 = CHARSETS["cp866"].decode(cp866).array;


Every codec accepts range of some chars and returns range of ubytes. Every decoder accepts range of ubytes  and returns range of dchars. This library is not cover UTF codecs, and multibytes national code pages.