%EXPORT_TAGS = (mime => [qw(parse create_mime_message getmailbody)],
rfc1522 => [qw(decode_rfc1522 encode_rfc1522)],
- utf8 => [qw(convert_to_utf8)],
);
@EXPORT_OK=();
Exporter::export_ok_tags(keys %EXPORT_TAGS);
use POSIX qw(strftime);
use List::MoreUtils qw(apply);
+# for convert_to_utf8
+use Debbugs::UTF8 qw(convert_to_utf8 encode_utf8_safely);
+
# for decode_rfc1522
use MIME::WordDecoder qw();
use Encode qw(decode encode encode_utf8 decode_utf8 is_utf8);
if ($type eq 'text/plain' or
($type =~ m#text/?# and $type ne 'text/html') or
$type eq 'application/pgp') {
- return $entity->bodyhandle;
+ return $entity;
} elsif ($type eq 'multipart/alternative') {
# RFC 2046 says we should use the last part we recognize.
for my $part (reverse $entity->parts) {
@headerlines = @{$entity->head->header};
chomp @headerlines;
- my $entity_body = getmailbody($entity);
- @bodylines = $entity_body ? $entity_body->as_lines() : ();
+ my $entity_body = getmailbody($entity);
+ my $entity_body_handle;
+ my $charset;
+ if (defined $entity_body) {
+ $entity_body_handle = $entity_body->bodyhandle();
+ $charset = $entity_body->head()->mime_attr('content-type.charset');
+ }
+ @bodylines = $entity_body_handle ? $entity_body_handle->as_lines() : ();
+ @bodylines = map {convert_to_utf8($_,$charset)} @bodylines;
chomp @bodylines;
} else {
# Legacy pre-MIME code, kept around in case MIME::Parser fails.
# MIME::Entity is stupid, and doesn't rfc1522 encode its headers, so we do it for it.
my $msg = MIME::Entity->build('Content-Type' => 'text/plain; charset=utf-8',
'Encoding' => 'quoted-printable',
- (map{encode_rfc1522($_)} @{$headers}),
- Data => $body
+ (map{encode_rfc1522(encode_utf8_safely($_))} @{$headers}),
+ Data => encode_utf8_safely($body),
);
# Attach the attachments
}
-# Bug #61342 et al.
-
-sub convert_to_utf8 {
- my ($data, $charset) = @_;
- # raw data just gets returned (that's the charset WordDecorder
- # uses when it doesn't know what to do)
- return $data if $charset eq 'raw' or is_utf8($data,1);
- my $result;
- eval {
- $result = decode($charset,$data);
- };
- if ($@) {
- warn "Unable to decode charset; '$charset' and '$data': $@";
- return $data;
- }
- return $result;
-}
=head2 decode_rfc1522
# handle being passed undef properly
return undef if not defined $rawstr;
- if (is_utf8($rawstr)) {
- $rawstr= encode_utf8($rawstr);
- }
+ # convert to octets if is_utf8 is set
+ $rawstr= encode_utf8_safely($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. We also consider parenthesis and "