Coding systems documentation

Coding systems documentation

[Richard M Stallman]

    +@c I think this paragraph is no longer correct.
    +@ignore
       Most coding systems specify a particular character code for
     conversion, but some of them leave the choice unspecified---to be chosen
     heuristically for each file, based on the data.
    +@end ignore

I think these still exist.  For example, there are `undecided' and friends.

Re: Coding systems documentation

[Eli Zaretskii]

> From: Richard M Stallman <rms@gnu.org>
> Date: Sun, 30 Nov 2008 11:43:56 -0500
> 
>     +@c I think this paragraph is no longer correct.
>     +@ignore
>        Most coding systems specify a particular character code for
>      conversion, but some of them leave the choice unspecified---to be chosen
>      heuristically for each file, based on the data.
>     +@end ignore
> 
> I think these still exist.  For example, there are `undecided' and friends.

Is this only about undecided? or are there other examples?

If it's only about undecided, I'd like to rewrite this text to speak
explicitly about undecided.  Then it won't sound so mysteriously.

Re: Coding systems documentation

[Kenichi Handa]

In article <uoczwlx0q.fsf@gnu.org>, Eli Zaretskii <eliz@gnu.org> writes:

> > From: Richard M Stallman <rms@gnu.org>
> > Date: Sun, 30 Nov 2008 11:43:56 -0500
> > 
> >     +@c I think this paragraph is no longer correct.
> >     +@ignore
> >        Most coding systems specify a particular character code for
> >      conversion, but some of them leave the choice unspecified---to be chosen
> >      heuristically for each file, based on the data.
> >     +@end ignore
> > 
> > I think these still exist.  For example, there are `undecided' and friends.

> Is this only about undecided? or are there other examples?

All coding systems that don't have -unix, -dos, and -mac at
the tail leaves the choice of eol-format unspecified.  But,
if you are going to mention about eol-format at the
different place, yes, `undecided' is the only coding that
doesn't specify how text is encode.

> If it's only about undecided, I'd like to rewrite this text to speak
> explicitly about undecided.  Then it won't sound so mysteriously.

It seems to be a good idea.

By the way, "specify a particular character code for
conversion" is a little bit strange.  "specify a particular
conversion rule between an encoded byte sequence and a
character sequence." is more accurate (and I think clearer).

---
Kenichi Handa
handa@ni.aist.go.jp

Re: Coding systems documentation

[Eli Zaretskii]

> From: Kenichi Handa <handa@m17n.org>
> CC: rms@gnu.org, emacs-devel@gnu.org
> Date: Mon, 01 Dec 2008 10:30:49 +0900
> 
> > >     +@c I think this paragraph is no longer correct.
> > >     +@ignore
> > >        Most coding systems specify a particular character code for
> > >      conversion, but some of them leave the choice unspecified---to be chosen
> > >      heuristically for each file, based on the data.
> > >     +@end ignore
> > > 
> > > I think these still exist.  For example, there are `undecided' and friends.
> 
> > Is this only about undecided? or are there other examples?
> 
> All coding systems that don't have -unix, -dos, and -mac at
> the tail leaves the choice of eol-format unspecified.

Yes, but the above says "character code", so it's about text
conversion, not eol conversion.

> By the way, "specify a particular character code for
> conversion" is a little bit strange.  "specify a particular
> conversion rule between an encoded byte sequence and a
> character sequence." is more accurate (and I think clearer).

Right.

Re: Coding systems documentation

[Kenichi Handa]

In article <ufxl8leke.fsf@gnu.org>, Eli Zaretskii <eliz@gnu.org> writes:

> > From: Kenichi Handa <handa@m17n.org>
> > CC: rms@gnu.org, emacs-devel@gnu.org
> > Date: Mon, 01 Dec 2008 10:30:49 +0900
> > 
> > > >     +@c I think this paragraph is no longer correct.
> > > >     +@ignore
> > > >        Most coding systems specify a particular character code for
> > > >      conversion, but some of them leave the choice unspecified---to be chosen
> > > >      heuristically for each file, based on the data.
> > > >     +@end ignore
> > > > 
> > > > I think these still exist.  For example, there are `undecided' and friends.
> > 
> > > Is this only about undecided? or are there other examples?
> > 
> > All coding systems that don't have -unix, -dos, and -mac at
> > the tail leaves the choice of eol-format unspecified.

> Yes, but the above says "character code", so it's about text
> conversion, not eol conversion.

LF and CR are also character code.

> > By the way, "specify a particular character code for
> > conversion" is a little bit strange.  "specify a particular
> > conversion rule between an encoded byte sequence and a
> > character sequence." is more accurate (and I think clearer).

> Right.

Then, the distinction of text conversion and eol conversion
should be stated beforehand.

---
Kenichi Handa
handa@ni.aist.go.jp

Re: Coding systems documentation

[Richard M Stallman]

    > I think these still exist.  For example, there are `undecided' and friends.

    Is this only about undecided? or are there other examples?

I think it is only that, plus the issue of line-end conversion.