Re: libcodecs(3), take 3

[SODA Noriyuki <soda%yuruyuru.net@localhost>]

Hi Al,

I forgot to write one thing.

If the following change is done, it may be OK to import libcodecs into
the tree, and fix other things later:
- changing the name "ascii2ebcdic" to "iconv(ASCII,EBCDIC-CP-US)"
- changing the name "ebcdic2ascii" to "iconv(EBCDIC-CP-US,ASCII)"
- changing "to_unicode" to "iconv(UTF-8,UTF-16LE)"
- changing "to_utf8" to "iconv(UTF-16LE,UTF-8)"

I think the followings must be done until any official release,
but that may not be trivial for people who are not so familiar with
iconv(3)/wctran(3) APIs, so it may be better to import libcodec at
first, and let people who familiar with the APIs change the usage.:
- libcodecs(3) should internally call iconv(3).
- wide character conversion should be done before calling towctrans(3).
-- 
soda