return $data if $charset eq 'raw' or is_utf8($data,1);
my $result;
eval {
- # this encode/decode madness is to make sure that the data
- # really is valid utf8 and that the is_utf8 flag is off.
- $result = encode("utf8",decode($charset,$data))
+ $result = decode($charset,$data);
};
if ($@) {
warn "Unable to decode charset; '$charset' and '$data': $@";
# handle being passed undef properly
return undef if not defined $rawstr;
+ if (is_utf8($rawstr)) {
+ $rawstr= encode_utf8($rawstr);
+ }
# We process words in reverse so we can preserve spacing between
# encoded words. This regex splits on word|nonword boundaries and
- # nonword|nonword boundaries.
- my @words = reverse split /(?:(?<=[\s\n])|(?=[\s\n]))/m, $rawstr;
+ # nonword|nonword boundaries. We also consider parenthesis and "
+ # to be nonwords to avoid escaping them in comments in violation
+ # of RFC1522
+ my @words = reverse split /(?:(?<=[\s\n\)\(\"])|(?=[\s\n\)\(\"]))/m, $rawstr;
my $previous_word_encoded = 0;
my $string = '';