Re: Bug 130397 (Was: Emacs - Ispell problem with i[no]german dictionary)

[Agustin Martin <agustin.martin@hispalinux.es>]

On Tue, Jan 04, 2005 at 09:50:33PM +0900, Kenichi Handa wrote:

> Hmmm, then how about the attached patch to the latest CVS
> emacs?  With that, all equivalent charaters (e.g a-grave in
> all laitn-X) should be handled well.  This patch will be
> applicable also to Emacs 21.3 but not yet tested in that
> version.
> 

Hi,

Thanks for the patch, and sorry for not being very responsive these days. I
can hardly keep up to date with my mail.

Your patch applies cleanly to ispell.el shipped with dictionaries-common
(just some minor shift), since it is 3.6 as well as for CVS emacs.

First noting that I still get ispell misalignement errors when spellchecking
an utf8 file with an iso-8859-15 dict for chars not in the iso-8859-1 set
(oe-char).

I also had some problems when making it work with the different emacsen
flavours. For sid emacs21 (21.3+1-8) works with some limitations, that seem
somewhat similar to those for my dirty hack. However for older versions
(e.g., woody emacs21) ucs-mule-8859-to-mule-unicode seems not avilable, as
well as for emacs20, so I get

Symbol's value as variable is void: ucs-mule-8859-to-mule-unicode

For xemacs, besides the known problems with buffer-file-coding-system I also
get the error message

Unrecognized char table type: ispell-unified-chars-table

Note that ispell.el at dictionaries-common package must work with all the
shipped emacsen flavours (unless the problem is due to a bug in the emacsen
flavour), currently emacs21 and xemacs21-{mule,nomule...} and preferrably to
not make things worse for previous flavours (emacs20, no need of going back
more, even if I sometimes test emacs19), so we need to be very conservative.
Some checks are possible, however, but I am rather undecided about that.

Thanks a lot for your feedback,

Cheers,

-- 
Agustin