bug#1187: 23.0.60; Cannot read vline.el - invalid read syntax

[Kenichi Handa <handa@m17n.org>]

In article <uy70jilch.fsf@gnu.org>, Eli Zaretskii <eliz@gnu.org> writes:

> It's vline.el, and you can download it from here:

>   http://www.emacswiki.org/cgi-bin/wiki/vline.el

> (It is better to download, to avoid any possible transformations of
> the file in mail transfer.)

Ok, I found what is the problem.  In Emacs 22, iso-8859-1 is
a coding of type `iso-2022', but in Emacs 23, it's of type
`charset', and I forgot to handle latin-extra-code-table
(which has nil for the byte 0x80) in the detector of that
kind of coding system.  As I've just installed a fix, the
latest code should detect vline.el correctly as utf-8.

But, the reason why it's detected as utf-8 is because the
file contains a byte 0x80.  If a file doesn't contain a byte
in the range 0x80..0x9F and nil in latin-extra-code-table,
it's impossible to distinguish iso-latin-1 from utf-8.

So, in general, it's a good idea to add coding-tag: utf-8
for utf-8 files.

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