From: Martin Duerst <duerst@...>
Date: 2009-02-08T16:31:31+09:00
Subject: [ruby-core:21927] Re: mule-utf-8

I think for the reported issue, the best thing is to fix
  -*- coding: mule-utf-8 -*-
to -*- coding: utf-8 -*-

As for the definition of mule-utf-8, I'm also not familiar with it,
but the best information I found was:
http://www.koders.com/lisp/fid3DDD5163F8F196DF87FA2396D27C77E592DC2C3C.aspx?s=sort

;; The coding-system `mule-utf-8' basically supports encoding/decoding
;; of the following character sets to and from UTF-8:
;;
;;   ascii
;;   eight-bit-control
;;   latin-iso8859-1
;;   mule-unicode-0100-24ff
;;   mule-unicode-2500-33ff
;;   mule-unicode-e000-ffff
;;
;; On decoding, Unicode characters that do not fit into the above
;; character sets are handled as `eight-bit-control' or
;; `eight-bit-graphic' characters to retain the information about the
;; original byte sequence and text properties record the corresponding
;; unicode.

So mule-utf-8 seems to be used to indicate a subset of UTF-8,
or UTF-8 with some special mule-internal treatment for some
specific character range (U+3400-DFFF, which includes CJK ideographs).

One of my guesses is that this special treatement was introduced
to avoid deciding which of C/J/K the ideographs should go to.
As Ruby supports UTF-8 fully, without the need to decide whether
an ideograph is Chinese, Japanese, or Korean,..., that would mean
that we can handle mule-utf-8 simply as UTF-8. But because
mule-utf-8 in that case is simply an artefact of some mule-internal
details, I'd be rather reluctant to support such a label at all.

Regards,    Martin.


At 02:11 09/02/07, you wrote:
>I'm also not familiar with this encoding, I was surprised as well.
>While the link below is not authoritative information, it may hint, that 
>this is an alias for utf-8.
>http://groups.google.com/group/gnu.emacs.help/browse_thread/thread/e7dc49d8ec4b937f?pli=1
>
>
>My original question was, whenever the support for this encoding wasn't 
>included by mistake (since I've seen some emacs related encodings file), or not.
>
>--Roman
>
>
>----- Original Message ----
>From: "NARUSE, Yui" <naruse@airemix.jp>
>To: ruby-core@ruby-lang.org
>Sent: Friday, February 6, 2009 6:23:06 PM
>Subject: [ruby-core:21892] Re: mule-utf-8
>
>Hi,
>
>Roman Shterenzon wrote:
>> When I attempted to run some rails application with Ruby 1.9, the
>> following exception was raised:
>> 
>> unknown encoding name: mule-utf-8
>> 
>> The offending file, so it turns out, was in the
>> restful-authentication plugin, which had this coding specified by: #
>> -*- coding: mule-utf-8 -*-
>> 
>> The question is, whenever this is a Ruby bug, and this encoding
>> should be supported (there's a file called enc/emacs_mule.c, which
>> might be related).
>
>You know ruby can't support all the encodings in the world.
>But we want to support many useful encodings.
>
>I didn't know mule-utf-8 before this time, and I know only the name.
>By some googled information, mule-utf-8 seems different from UTF-8.
>(if a encoding equals UTF-8, add the alias is easy)
>If ruby should support the encoding, we want more information about its
>byte structure.
>
>
>-- NARUSE, Yui  <naruse@airemix.jp>
>
>
>      


#-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst@it.aoyama.ac.jp