1 # This module is part of debbugs, and is released
2 # under the terms of the GPL version 2, or any later
3 # version at your option.
4 # See the file README and COPYING for more information.
6 # [Other people have contributed to this file; their copyrights should
8 # Copyright 2006 by Don Armstrong <don@donarmstrong.com>.
11 package Debbugs::MIME;
15 use base qw(Exporter);
16 use vars qw($VERSION @EXPORT_OK);
21 @EXPORT_OK = qw(parse decode_rfc1522 encode_rfc1522 convert_to_utf8 create_mime_message);
28 use MIME::WordDecoder qw();
29 use Encode qw(decode encode encode_utf8 decode_utf8 is_utf8);
38 my $type = $entity->effective_type;
39 if ($type eq 'text/plain' or
40 ($type =~ m#text/# and $type ne 'text/html') or
41 $type eq 'application/pgp') {
42 return $entity->bodyhandle;
43 } elsif ($type eq 'multipart/alternative') {
44 # RFC 2046 says we should use the last part we recognize.
45 for my $part (reverse $entity->parts) {
46 my $ret = getmailbody($part);
50 # For other multipart types, we just pretend they're
51 # multipart/mixed and run through in order.
52 for my $part ($entity->parts) {
53 my $ret = getmailbody($part);
62 # header and decoded body respectively
63 my (@headerlines, @bodylines);
65 my $parser = new MIME::Parser;
66 mkdir "mime.tmp.$$", 0777;
67 $parser->output_under("mime.tmp.$$");
68 my $entity = eval { $parser->parse_data($_[0]) };
70 if ($entity and $entity->head->tags) {
71 @headerlines = @{$entity->head->header};
74 my $entity_body = getmailbody($entity);
75 @bodylines = $entity_body ? $entity_body->as_lines() : ();
78 # Legacy pre-MIME code, kept around in case MIME::Parser fails.
79 my @msg = split /\n/, $_[0];
82 for ($i = 0; $i <= $#msg; ++$i) {
85 while ($msg[$i + 1] =~ /^\s/) {
87 $_ .= "\n" . $msg[$i];
89 push @headerlines, $_;
92 @bodylines = @msg[$i .. $#msg];
95 rmtree "mime.tmp.$$", 0, 1;
98 shift @bodylines while @bodylines and $bodylines[0] !~ /\S/;
100 # Strip off RFC2440-style PGP clearsigning.
101 if (@bodylines and $bodylines[0] =~ /^-----BEGIN PGP SIGNED/) {
102 shift @bodylines while @bodylines and length $bodylines[0];
103 shift @bodylines while @bodylines and $bodylines[0] !~ /\S/;
104 for my $findsig (0 .. $#bodylines) {
105 if ($bodylines[$findsig] =~ /^-----BEGIN PGP SIGNATURE/) {
106 $#bodylines = $findsig - 1;
110 map { s/^- // } @bodylines;
113 return { header => [@headerlines], body => [@bodylines]};
116 =head2 create_mime_message
118 create_mime_message([To=>'don@debian.org'],$body,[$attach1, $attach2]);
120 Creates a MIME encoded message with headers given by the first
121 argument, and a message given by the second.
123 Optional attachments can be specified in the third arrayref argument.
125 Headers are passed directly to MIME::Entity::build, the message is the
128 Each of the elements of the attachment arrayref is attached as an
129 rfc822 message if it is a scalar or an arrayref; otherwise if it is a
130 hashref, the contents are passed as an argument to
135 sub create_mime_message{
136 my ($headers,$body,$attachments) = @_;
137 $attachments = [] if not defined $attachments;
139 die "The first argument to create_mime_message must be an arrayref" unless ref($headers) eq 'ARRAY';
140 die "The third argument to create_mime_message must be an arrayref" unless ref($attachments) eq 'ARRAY';
143 # MIME::Entity is stupid, and doesn't rfc1522 encode its headers, so we do it for it.
144 my $msg = MIME::Entity->build('Content-Type' => 'text/plain; charset=utf-8',
145 'Encoding' => 'quoted-printable',
146 (map{encode_rfc1522($_)} @{$headers}),
150 # Attach the attachments
151 for my $attachment (@{$attachments}) {
152 if (ref($attachment) eq 'HASH') {
153 $msg->attach(%{$attachment});
156 # This is *craptacular*, but because various MTAs
157 # (sendmail and exim4, at least) appear to eat From
158 # lines in message/rfc822 attachments, we need eat
159 # the entire From line ourselves so the MTA doesn't
160 # leave \n detrius around.
161 if (ref($attachment) eq 'ARRAY' and $attachment->[1] =~ /^From /) {
162 # make a copy so that we don't screw up anything
163 # that is expecting this arrayref to stay constant
164 $attachment = [@{$attachment}];
165 # remove the from line
166 splice @$attachment, 1, 1;
168 elsif (not ref($attachment)) {
169 # It's a scalar; remove the from line
170 $attachment =~ s/^(Received:[^\n]+\n)(From [^\n]+\n)/$1/s;
172 $msg->attach(Type => 'message/rfc822',
178 return $msg->as_string;
184 sub convert_to_utf8 {
185 my ($data, $charset) = @_;
186 # raw data just gets returned (that's the charset WordDecorder
187 # uses when it doesn't know what to do)
188 return $data if $charset eq 'raw' or is_utf8($data,1);
191 # this encode/decode madness is to make sure that the data
192 # really is valid utf8 and that the is_utf8 flag is off.
193 $result = encode("utf8",decode($charset,$data))
196 warn "Unable to decode charset; '$charset' and '$data': $@";
203 =head2 decode_rfc1522
205 decode_rfc1522('=?iso-8859-1?Q?D=F6n_Armstr=F3ng?= <don@donarmstrong.com>')
207 Turn RFC-1522 names into the UTF-8 equivalent.
212 # Set up the default RFC1522 decoder, which turns all charsets that
213 # are supported into the appropriate UTF-8 charset.
214 MIME::WordDecoder->default(new MIME::WordDecoder(
215 ['*' => \&convert_to_utf8,
219 sub decode_rfc1522 ($)
223 # this is craptacular, but leading space is hacked off by unmime.
225 my $leading_space = '';
226 $leading_space = $1 if $string =~ s/^(\s+)//;
227 # unmime calls the default MIME::WordDecoder handler set up at
228 # initialization time.
229 return $leading_space . MIME::WordDecoder::unmime($string);
232 =head2 encode_rfc1522
234 encode_rfc1522('Dön Armströng <don@donarmstrong.com>')
236 Encodes headers according to the RFC1522 standard by calling
237 MIME::Words::encode_mimeword on distinct words as appropriate.
241 # We cannot use MIME::Words::encode_mimewords because that function
242 # does not handle spaces properly at all.
244 sub encode_rfc1522 ($) {
247 # We process words in reverse so we can preserve spacing between
248 # encoded words. This regex splits on word|nonword boundaries and
249 # nonword|nonword boundaries.
250 my @words = reverse split /(?:(?<=[\s\n])|(?=[\s\n]))/m, $rawstr;
252 my $previous_word_encoded = 0;
254 for my $word (@words) {
255 if ($word !~ m#[\x00-\x1F\x7F-\xFF]#o and $word ne ' ') {
256 $string = $word.$string;
257 $previous_word_encoded=0;
259 elsif ($word =~ /^[\s\n]$/) {
260 $string = $word.$string;
261 $previous_word_encoded = 0 if $word eq "\n";
264 my $encoded = MIME::Words::encode_mimeword($word, 'q', 'UTF-8');
265 # RFC 1522 mandates that segments be at most 76 characters
266 # long. If that's the case, we split the word up into 10
267 # character pieces and encode it. We must use the Encode
268 # magic here to avoid breaking on bit boundaries here.
269 if (length $encoded > 75) {
270 # Turn utf8 into the internal perl representation
271 # so . is a character, not a byte.
272 my $tempstr = decode_utf8($word,Encode::FB_DEFAULT);
274 # Strip it into 10 character long segments, and encode
276 # XXX It's possible that these segments are > 76 characters
277 while ($tempstr =~ s/(.{1,10})$//) {
278 # turn the character back into the utf8 representation.
279 my $tempword = encode_utf8($1);
280 # It may actually be better to eventually use
281 # the base64 encoding here, but I'm not sure
282 # if that's as widely supported as quoted
284 unshift @encoded, MIME::Words::encode_mimeword($tempword,'q','UTF-8');
286 $encoded = join(" ",@encoded);
287 # If the previous word was encoded, we must
288 # include a trailing _ that gets encoded as a
290 $encoded =~ s/\?\=$/_\?\=/ if $previous_word_encoded;
291 $string = $encoded.$string;
294 # If the previous word was encoded, we must
295 # include a trailing _ that gets encoded as a
297 $encoded =~ s/\?\=$/_\?\=/ if $previous_word_encoded;
298 $string = $encoded.$string;
300 $previous_word_encoded = 1;