> On Thursday 16 March 2006 12:14, Agustin Martin wrote:
> > When ispell reads a .tex file both deformatter and charset are set to tex
> > giving the behavior described in #61956. To deal with this you just need to
> > explicitely specify the charset, e.g., for spanish
> >
> > $ ispell -dspanish -Tlatin1 test.tex
> >
> > works as expected. For italian or french something like
> >
> > $ ispell -ditalian -Tlist test.tex
> >
> > should work.
I wrote too fast, spanish has tex as default deformatter for latin1, so
worked well, but I now expect problems for other languages, we still have a
problem.
On Thu, Mar 16, 2006 at 02:02:15PM +0100, David Schmitt wrote:
> I'd expect similar problems with HTML (which can contain various encodings
> too). A real fix would be to separate charset handling and formatting
> completely or to extend the tex formatter to recognise the
> \usepackage[charset]{inputenc} declaration.
Agreed that they should be separated.
The problem is that ispell declarations for a format (in the .aff file) are
something like (taken from italian.aff)
defstringtype "list" "nroff" ".list" ".txt"
char equivalences
where list is the format name, nroff the deformatter and the rest of the
line a list of suffixes where that format is to be used. Any of "-(t|n|h)"
or things like "-Tlist" will pull the entire set, not only the charset or
the deformatter.
Workaround is that ispell dicts provide both tex and tex8 for tex deformatter
and 8bit charset. I agree, this is ugly.
Looking at ispell 3.3 series I see that a new option "-F" is added to
support external deformatters, that will be default sometime.
"The -F switch is the preferred way to deformat files for ispell, and
eventually will become the only way."
but I saw no external deformatters for tex.
Seems that these bug reports will stay open for long.
--
Agustin
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
