X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=Debbugs%2FMIME.pm;h=481be7bf9573f758ff331292b8f5a61fe66b75a6;hb=9501bf7323cb40d0365f589fad6068dba4c48a45;hp=59ffda489b47c282d5c23d850177e07f1501b434;hpb=adad9c647a843eb1d491aaa8e4493d847b5b98fe;p=debbugs.git diff --git a/Debbugs/MIME.pm b/Debbugs/MIME.pm index 59ffda4..481be7b 100644 --- a/Debbugs/MIME.pm +++ b/Debbugs/MIME.pm @@ -10,20 +10,51 @@ package Debbugs::MIME; +=head1 NAME + +Debbugs::MIME -- Mime handling routines for debbugs + +=head1 SYNOPSIS + + use Debbugs::MIME qw(parse decode_rfc1522); + +=head1 DESCRIPTION + + +=head1 BUGS + +None known. + +=cut + +use warnings; use strict; use base qw(Exporter); -use vars qw($VERSION @EXPORT_OK); +use vars qw($DEBUG $VERSION @EXPORT_OK %EXPORT_TAGS @EXPORT); BEGIN { $VERSION = 1.00; + $DEBUG = 0 unless defined $DEBUG; + + @EXPORT = (); - @EXPORT_OK = qw(parse decode_rfc1522 encode_rfc1522 convert_to_utf8 create_mime_message getmailbody); + %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); + $EXPORT_TAGS{all} = [@EXPORT_OK]; } use File::Path; +use File::Temp qw(); use MIME::Parser; +use POSIX qw(strftime); +use List::MoreUtils qw(apply); + # for decode_rfc1522 use MIME::WordDecoder qw(); use Encode qw(decode encode encode_utf8 decode_utf8 is_utf8); @@ -36,7 +67,7 @@ sub getmailbody my $entity = shift; my $type = $entity->effective_type; if ($type eq 'text/plain' or - ($type =~ m#text/# and $type ne 'text/html') or + ($type =~ m#text/?# and $type ne 'text/html') or $type eq 'application/pgp') { return $entity->bodyhandle; } elsif ($type eq 'multipart/alternative') { @@ -56,14 +87,14 @@ sub getmailbody return undef; } -sub parse ($) +sub parse { # header and decoded body respectively my (@headerlines, @bodylines); - my $parser = new MIME::Parser; - mkdir "mime.tmp.$$", 0777; - $parser->output_under("mime.tmp.$$"); + my $parser = MIME::Parser->new(); + my $tempdir = File::Temp::tempdir(); + $parser->output_under($tempdir); my $entity = eval { $parser->parse_data($_[0]) }; if ($entity and $entity->head->tags) { @@ -91,14 +122,19 @@ sub parse ($) @bodylines = @msg[$i .. $#msg]; } - rmtree "mime.tmp.$$", 0, 1; + rmtree $tempdir, 0, 1; # Remove blank lines. shift @bodylines while @bodylines and $bodylines[0] !~ /\S/; # Strip off RFC2440-style PGP clearsigning. if (@bodylines and $bodylines[0] =~ /^-----BEGIN PGP SIGNED/) { - shift @bodylines while @bodylines and length $bodylines[0]; + shift @bodylines while @bodylines and + length $bodylines[0] and + # we currently don't strip \r; handle this for the + # time being, though eventually it should be stripped + # too, I think. [See #565981] + $bodylines[0] ne "\r"; shift @bodylines while @bodylines and $bodylines[0] !~ /\S/; for my $findsig (0 .. $#bodylines) { if ($bodylines[$findsig] =~ /^-----BEGIN PGP SIGNATURE/) { @@ -114,13 +150,17 @@ sub parse ($) =head2 create_mime_message - create_mime_message([To=>'don@debian.org'],$body,[$attach1, $attach2]); + create_mime_message([To=>'don@debian.org'],$body,[$attach1, $attach2],$include_date); Creates a MIME encoded message with headers given by the first argument, and a message given by the second. Optional attachments can be specified in the third arrayref argument. +Whether to include the date in the header is the final argument; it +defaults to true, setting the Date header if one is not already +present. + Headers are passed directly to MIME::Entity::build, the message is the first attachment. @@ -132,12 +172,23 @@ MIME::Entity::attach =cut sub create_mime_message{ - my ($headers,$body,$attachments) = @_; + my ($headers,$body,$attachments,$include_date) = @_; $attachments = [] if not defined $attachments; + $include_date = 1 if not defined $include_date; die "The first argument to create_mime_message must be an arrayref" unless ref($headers) eq 'ARRAY'; die "The third argument to create_mime_message must be an arrayref" unless ref($attachments) eq 'ARRAY'; + if ($include_date) { + my %headers = apply {lc($_)} @{$headers}; + if (not exists $headers{date}) { + push @{$headers}, + ('Date', + strftime("%a, %d %b %Y %H:%M:%S +0000",gmtime) + ); + } + } + # Build the message # 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', @@ -187,9 +238,7 @@ sub convert_to_utf8 { 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': $@"; @@ -215,8 +264,7 @@ BEGIN { ])); } -sub decode_rfc1522 ($) -{ +sub decode_rfc1522 { my ($string) = @_; # this is craptacular, but leading space is hacked off by unmime. @@ -240,13 +288,20 @@ MIME::Words::encode_mimeword on distinct words as appropriate. # We cannot use MIME::Words::encode_mimewords because that function # does not handle spaces properly at all. -sub encode_rfc1522 ($) { +sub encode_rfc1522 { my ($rawstr) = @_; + # 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 $string = ''; @@ -268,7 +323,7 @@ sub encode_rfc1522 ($) { if (length $encoded > 75) { # Turn utf8 into the internal perl representation # so . is a character, not a byte. - my $tempstr = decode_utf8($word,Encode::FB_DEFAULT); + my $tempstr = is_utf8($word)?$word:decode_utf8($word,Encode::FB_DEFAULT); my @encoded; # Strip it into 10 character long segments, and encode # the segments