>> >> $ groff -V -s -Kcp1251 -t -Tutf8
>> >> soelim | preconv -ecp1251 | tbl | troff -Tutf8 | grotty
>> >
>> > Not that you should ever do this, but if you were using, say,
>> > UTF-16, soelim would not be very effective unless preconv ran
>> > first.
>
> Could soelim be rewritten to use fgetwc() and wide characters?
> Standard ASCII would just pass through.
This doesn't help. `soelim' must recognize the string ".so", but in
UTF-16, this is represented as
"\0.\0s\0o"
and without knowing the encoding you can't find that in a document
easily. You are entering very muddy waters if you start to guess
encodings...
Werner
