> What do people think about this attached patch?
>
> Andy
If you find the word
"coding" by scanning 8-bit char by 8-bit char, it can't
be UTF-16, since that would be more like
"c o d i n g :" with nulls interspersed.
While rather unlikely, it is a theoretical possibility
that a doc in encodings like ISO-8859-2 through 8859-5
could begin with 0xff 0xfe or 0xfe 0xff. They are
valid characters.
So if there is a "coding:" line in the doc, I think it
should nullify giving precedence to a UTF-16 BOM.
-Mike Gran
