X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FMbox.pm;h=04c86cc196515e345d36491b200a1d29f3e4bd17;hb=15eb65ffd59c7cf5cce67c6b7621f63194fd1083;hp=fcb266939e1707e39d3e82b70751cd3e600793a2;hpb=7e3c392a55686d2ffe8611483e80a3ccc19e95f2;p=public-inbox.git diff --git a/lib/PublicInbox/Mbox.pm b/lib/PublicInbox/Mbox.pm index fcb26693..04c86cc1 100644 --- a/lib/PublicInbox/Mbox.pm +++ b/lib/PublicInbox/Mbox.pm @@ -1,74 +1,122 @@ -# Copyright (C) 2015, all contributors -# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt) -# Streaming interface for formatting messages as an mbox +# Copyright (C) 2015-2018 all contributors +# License: AGPL-3.0+ + +# Streaming interface for formatting messages as an mboxrd. +# Used by the web interface package PublicInbox::Mbox; use strict; use warnings; -use PublicInbox::MID qw/mid_compressed mid2path/; +use PublicInbox::MID qw/mid_clean mid_escape/; +use PublicInbox::Hval qw/to_filename/; +use Email::Simple; +use Email::MIME::Encode; + +sub subject_fn ($) { + my ($simple) = @_; + my $fn = $simple->header('Subject'); + return 'no-subject' unless defined($fn); + + # no need for full Email::MIME, here + if ($fn =~ /=\?/) { + eval { $fn = Encode::decode('MIME-Header', $fn) }; + $fn = 'no-subject' if $@; + } + $fn =~ s/^re:\s+//i; + $fn = to_filename($fn); + $fn eq '' ? 'no-subject' : $fn; +} -sub thread_mbox { - my ($ctx, $srch) = @_; - sub { - my ($response) = @_; # Plack callback - emit_mbox($response, $ctx, $srch); +sub emit1 { + my ($ctx, $msg) = @_; + $msg = Email::Simple->new($msg); + my $fn = subject_fn($msg); + my @hdr = ('Content-Type'); + if ($ctx->{-inbox}->{obfuscate}) { + # obfuscation is stupid, but maybe scrapers are, too... + push @hdr, 'application/mbox'; + $fn .= '.mbox'; + } else { + push @hdr, 'text/plain'; + $fn .= '.txt'; } + push @hdr, 'Content-Disposition', "inline; filename=$fn"; + + # single message should be easily renderable in browsers, + # unless obfuscation is enabled :< + [ 200, \@hdr, [ msg_str($ctx, $msg) ] ] } -sub emit_msg { - my ($fh, $simple) = @_; # Email::Simple object +sub msg_str { + my ($ctx, $simple) = @_; # Email::Simple object + my $header_obj = $simple->header_obj; # drop potentially confusing headers, ssoma already should've dropped # Lines and Content-Length - foreach my $d (qw(Lines Content-Length Status)) { - $simple->header_set($d); + foreach my $d (qw(Lines Bytes Content-Length Status)) { + $header_obj->header_set($d); } - - my $buf = $simple->header_obj->as_string; - unless ($buf =~ /\AFrom /) { - $fh->write("From a\@a Thu Jan 1 00:00:00 1970\n"); + my $ibx = $ctx->{-inbox}; + my $base = $ibx->base_url($ctx->{env}); + my $mid = mid_clean($header_obj->header('Message-ID')); + $mid = mid_escape($mid); + my @append = ( + 'Archived-At', "<$base$mid/>", + 'List-Archive', "<$base>", + 'List-Post', "{-primary_address}>", + ); + my $crlf = $simple->crlf; + my $buf = "From mboxrd\@z Thu Jan 1 00:00:00 1970\n" . + $header_obj->as_string; + for (my $i = 0; $i < @append; $i += 2) { + my $k = $append[$i]; + my $v = $append[$i + 1]; + my @v = $header_obj->header($k); + foreach (@v) { + if ($v eq $_) { + $v = undef; + last; + } + } + $buf .= "$k: $v$crlf" if defined $v; } - $fh->write($buf .= $simple->crlf); + $buf .= $crlf; + + # mboxrd quoting style + # ref: http://www.qmail.org/man/man5/mbox.html + my $body = $simple->body; + $body =~ s/^(>*From )/>$1/gm; + $buf .= $body; + $buf .= "\n"; +} - $buf = $simple->body; - $simple->body_set(''); - $buf =~ s/^(From )/>$1/gm; - $buf .= "\n" unless $buf =~ /\n\z/s; +sub thread_mbox { + my ($ctx, $srch, $sfx) = @_; + eval { require IO::Compress::Gzip }; + return sub { need_gzip(@_) } if $@; - $fh->write($buf); + my $cb = sub { $srch->get_thread($ctx->{mid}, @_) }; + PublicInbox::MboxGz->response($ctx, $cb); } -sub emit_mbox { - my ($response, $ctx, $srch) = @_; - eval { require IO::Compress::Gzip }; - return need_gzip($response) if $@; +sub emit_range { + my ($ctx, $range) = @_; - # http://www.iana.org/assignments/media-types/application/gzip - # http://www.iana.org/assignments/media-types/application/mbox - my $fh = $response->([200, ['Content-Type' => 'application/gzip']]); - $fh = PublicInbox::MboxGz->new($fh); - - require PublicInbox::GitCatFile; - require Email::Simple; - my $mid = mid_compressed($ctx->{mid}); - my $git = PublicInbox::GitCatFile->new($ctx->{git_dir}); - my %opts = (offset => 0); - my $nr; - do { - my $res = $srch->get_thread($mid, \%opts); - my $msgs = $res->{msgs}; - $nr = scalar @$msgs; - while (defined(my $smsg = shift @$msgs)) { - my $msg = eval { - my $p = 'HEAD:'.mid2path($smsg->mid); - Email::Simple->new($git->cat_file($p)); - }; - emit_msg($fh, $msg) if $msg; - } + my $query; + if ($range eq 'all') { # TODO: YYYY[-MM] + $query = ''; + } else { + return [404, [qw(Content-Type text/plain)], []]; + } + mbox_all($ctx, $query); +} - $opts{offset} += $nr; - } while ($nr > 0); +sub mbox_all { + my ($ctx, $query) = @_; - $fh->close; + eval { require IO::Compress::Gzip }; + return sub { need_gzip(@_) } if $@; + my $cb = sub { $ctx->{srch}->query($query, @_) }; + PublicInbox::MboxGz->response($ctx, $cb, 'results-'.$query); } sub need_gzip { @@ -80,45 +128,94 @@ The administrator needs to install the IO::Compress::Gzip Perl module to support gzipped mboxes. Return to index EOF + $fh->close; } 1; -# fh may not be a proper IO, so we wrap the write and close methods -# to prevent IO::Compress::Gzip from complaining package PublicInbox::MboxGz; use strict; use warnings; -use fields qw(gz fh buf); +use PublicInbox::Hval qw/to_filename/; sub new { - my ($class, $fh) = @_; - my $self = fields::new($class); - my $buf; - $self->{buf} = \$buf; - $self->{gz} = IO::Compress::Gzip->new(\$buf); - $self->{fh} = $fh; - $self; + my ($class, $ctx, $cb) = @_; + my $buf = ''; + bless { + buf => \$buf, + gz => IO::Compress::Gzip->new(\$buf, Time => 0), + cb => $cb, + ctx => $ctx, + msgs => [], + opts => { offset => 0 }, + }, $class; } -sub _flush_buf { - my ($self) = @_; - if (defined ${$self->{buf}}) { - $self->{fh}->write(${$self->{buf}}); - ${$self->{buf}} = undef; +sub response { + my ($class, $ctx, $cb, $fn) = @_; + my $body = $class->new($ctx, $cb); + # http://www.iana.org/assignments/media-types/application/gzip + $body->{hdr} = [ 'Content-Type', 'application/gzip' ]; + $body->{fn} = $fn; + # fill in Content-Disposition filename + my $hdr = $body->getline; + if ($body->{hdr}) { + return [ 404, ['Content-Type','text/plain'], + [ "No results found\n" ] ]; } + [ 200, $hdr, $body ]; } -sub write { - $_[0]->{gz}->write($_[1]); - _flush_buf($_[0]); +sub set_filename ($$) { + my ($fn, $msg) = @_; + return to_filename($fn) if defined($fn); + + PublicInbox::Mbox::subject_fn($msg); } -sub close { +# called by Plack::Util::foreach or similar +sub getline { my ($self) = @_; - $self->{gz}->close; - _flush_buf($self); - # do not actually close $fh + my $ctx = $self->{ctx} or return; + my $res; + my $ibx = $ctx->{-inbox}; + my $gz = $self->{gz}; + do { + # work on existing result set + while (defined(my $smsg = shift @{$self->{msgs}})) { + my $msg = eval { $ibx->msg_by_smsg($smsg) } or next; + $msg = Email::Simple->new($msg); + $gz->write(PublicInbox::Mbox::msg_str($ctx, $msg)); + + # use subject of first message as subject + if (my $hdr = delete $self->{hdr}) { + my $fn = set_filename($self->{fn}, $msg); + push @$hdr, 'Content-Disposition', + "inline; filename=$fn.mbox.gz"; + return $hdr; + } + my $bref = $self->{buf}; + if (length($$bref) >= 8192) { + my $ret = $$bref; # copy :< + ${$self->{buf}} = ''; + return $ret; + } + + # be fair to other clients on public-inbox-httpd: + return ''; + } + + # refill result set + $res = $self->{cb}->($self->{opts}); + $self->{msgs} = $res->{msgs}; + $res = scalar @{$self->{msgs}}; + $self->{opts}->{offset} += $res; + } while ($res); + $gz->close; + delete $self->{ctx}; + ${delete $self->{buf}}; } +sub close {} # noop + 1;