I noticed in when spellchecking non-roman languages that non-roman
words were silently and confusingly omitted from the dialog. It
turns out that we use UT_UCS_strcpy_to_char() to convert all words
to 8 bits. But that function doesn't insert '?' for nonconvertible
characters. I've ammended this.
If anyone can see a good reason why sometimes we would want these
characters silently omitted I can add a "bFallback" parameter to
the function. I can't see a reason.
Andrew Dunbar.
--
http://linguaphile.sourceforge.net
diff-ut_string.txt.gz
