> Make encode_from a bit more aggressive, by also catching a 'f' in
> the beginning of a line.
'Catch' is good but also to encode it is better. It may have
been too late (sorry for that) so I fixed it. I don't know
why the current source only escapes lines that actually
begin with 'from' or 'From' followed by some text. I think
that escaping 'from' or 'From' on a single line should be
done, too.