From: Jean-Hugues ROBERT <jean_hugues_robert@...>
Date: 2002-05-11T03:27:15+09:00
Subject: Re: Unicode in Ruby's Future?

Hello,

At 03:10 11/05/2002 +0900, you wrote:
>On 5/10/02 9:54 AM, "Yukihiro Matsumoto" <matz@ruby-lang.org> wrote:
>
> >  + string indexing is based on character, even for multi-byte
> >    characters.
> >  + regex routine understands multi-byte character strings.
> >  + user can define new character encoding scheme.
> >  + Ruby will shipped with support for ASCII, Latin-1, UTF-8, EUC-JP,
> >    SJIS, and probably more.
> >  + the goal of Ruby M17N is string manipulation without code
> >    conversion.
> >  + but there still be need for code conversion, but its spec is not
> >    fixed yet, probably it will use iconv.
> >
> > If you don't use multi-byte characters, there will be no
> > incompatibility.  If you do, string indexing for multi-byte character
> > string is not recommended; use regex match instead.
>
>Thanks to Matz and everyone who responded. I really like where Ruby is
>going.
>
>I infer from what is here that it's a poor idea for fake C structures like
>stacks with strings. We should use arrays for that kind of thing.

Hum... I was assuming that the current String would keep being byte based
and that there would be some kind of WideString when user needs multi-byte
characters. A more oo way would be to have AbstractString, ByteString and
xxxString. Might be overkill.

Yours,

Jean-Hugues