>>>> "dv" == Didier Verna
<verna(a)inf.enst.fr> writes:
dv> Martin Buchholz writes:
Martin> Conversion between internal and external string representation needs
Martin> to happen whenever data is passed between XEmacs and the `outside
Martin> world'.
dv> Yup. Then, I reformulate my question: when XEmacs retrieves a string
dv> from the outside world, say a pathname or the environment, how does it know
dv> the `outside' format? According to the locale it was started under? If
it's
dv> the case (so if XEmacs doesn't try to be clever), handling inconsistencies in
dv> the user's environment is problematic.
This has changed since I last worked on it. I think Tomohiko Morioka
is the last person who modified the XEmacs Mule code.
Try setting file-name-coding-system to something interesting, and see
what happens.
Martin> It sounds like the data in the lisp string may already have
Martin> been corrupt.
dv> Why? At the lisp level, everything looks correct. Even inside the
dv> debugger `call debug_print (current_dir)' produces a correct output.
Then the problem lies elsewhere.
Martin