X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FMbox.pm;h=05de6be1a678f75ac258b24233c9f52d8c7a585f;hb=87dca6d8d5988c5eb54019cca342450b0b7dd6b7;hp=dc41548b87ce68726dc41aa79bd571589a150641;hpb=d9732203aa53d1fd7b3ba5fae415d3ac003937a0;p=public-inbox.git diff --git a/lib/PublicInbox/Mbox.pm b/lib/PublicInbox/Mbox.pm index dc41548b..05de6be1 100644 --- a/lib/PublicInbox/Mbox.pm +++ b/lib/PublicInbox/Mbox.pm @@ -1,24 +1,98 @@ -# Copyright (C) 2015 all contributors -# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt) +# 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_clean/; -use URI::Escape qw/uri_escape_utf8/; -require Email::Simple; - -sub emit1 { - my ($ctx, $msg) = @_; - $msg = Email::Simple->new($msg); - # single message should be easily renderable in browsers - [200, ['Content-Type', 'text/plain'], [ msg_str($ctx, $msg)] ] +use PublicInbox::MID qw/mid_clean mid_escape/; +use PublicInbox::Hval qw/to_filename/; +use Email::Simple; +use Email::MIME::Encode; + +sub subject_fn ($) { + 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' : $fn; +} + +sub mb_stream { + my ($more) = @_; + bless $more, 'PublicInbox::Mbox'; +} + +# called by PSGI server as body response +sub getline { + my ($more) = @_; # self + my ($ctx, $head, $tail, $db, $cur) = @$more; + if ($cur) { + pop @$more; + return msg_str($ctx, $cur); + } + for (; !defined($cur) && $head != $tail; $head++) { + my $smsg = PublicInbox::SearchMsg->get($head, $db, $ctx->{mid}); + my $mref = $ctx->{-inbox}->msg_by_smsg($smsg) or next; + $cur = Email::Simple->new($mref); + $cur = msg_str($ctx, $cur); + } + $more->[1] = $head; + $cur; +} + +sub close {} # noop + +sub emit_raw { + my ($ctx) = @_; + my $mid = $ctx->{mid}; + my $ibx = $ctx->{-inbox}; + my $first; + my $more; + my ($head, $tail, $db); + my %seen; + if (my $srch = $ibx->search) { + $srch->retry_reopen(sub { + ($head, $tail, $db) = $srch->each_smsg_by_mid($mid); + for (; !defined($first) && $head != $tail; $head++) { + my @args = ($head, $db, $mid); + my $smsg = PublicInbox::SearchMsg->get(@args); + my $mref = $ibx->msg_by_smsg($smsg) or next; + $first = Email::Simple->new($mref); + } + if ($head != $tail) { + $more = [ $ctx, $head, $tail, $db, $first ]; + } + }); + } else { + my $mref = $ibx->msg_by_mid($mid) or return; + $first = Email::Simple->new($mref); + } + return unless defined $first; + my $fn = subject_fn($first); + my @hdr = ('Content-Type'); + if ($ibx->{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"; + [ 200, \@hdr, $more ? mb_stream($more) : [ msg_str($ctx, $first) ] ]; } sub msg_str { - my ($ctx, $simple) = @_; # Email::Simple object + my ($ctx, $simple, $mid) = @_; # Email::Simple object my $header_obj = $simple->header_obj; # drop potentially confusing headers, ssoma already should've dropped @@ -28,8 +102,8 @@ sub msg_str { } my $ibx = $ctx->{-inbox}; my $base = $ibx->base_url($ctx->{env}); - my $mid = mid_clean($header_obj->header('Message-ID')); - $mid = uri_escape_utf8($mid); + $mid = $ctx->{mid} unless defined $mid; + $mid = mid_escape($mid); my @append = ( 'Archived-At', "<$base$mid/>", 'List-Archive', "<$base>", @@ -66,27 +140,28 @@ sub thread_mbox { return sub { need_gzip(@_) } if $@; my $cb = sub { $srch->get_thread($ctx->{mid}, @_) }; - # http://www.iana.org/assignments/media-types/application/gzip - [200, ['Content-Type' => 'application/gzip'], - PublicInbox::MboxGz->new($ctx, $cb) ]; + PublicInbox::MboxGz->response($ctx, $cb); } sub emit_range { my ($ctx, $range) = @_; - eval { require IO::Compress::Gzip }; - return sub { need_gzip(@_) } if $@; my $query; if ($range eq 'all') { # TODO: YYYY[-MM] $query = ''; } else { return [404, [qw(Content-Type text/plain)], []]; } - my $cb = sub { $ctx->{srch}->query($query, @_) }; + mbox_all($ctx, $query); +} - # http://www.iana.org/assignments/media-types/application/gzip - [200, [qw(Content-Type application/gzip)], - PublicInbox::MboxGz->new($ctx, $cb) ]; +sub mbox_all { + my ($ctx, $query) = @_; + + 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 { @@ -106,6 +181,7 @@ EOF package PublicInbox::MboxGz; use strict; use warnings; +use PublicInbox::Hval qw/to_filename/; sub new { my ($class, $ctx, $cb) = @_; @@ -116,22 +192,49 @@ sub new { cb => $cb, ctx => $ctx, msgs => [], - opts => { asc => 1, offset => 0 }, + opts => { offset => 0 }, }, $class; } +sub response { + my ($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; + my $hdr = $body->getline; # fill in Content-Disposition filename + [ 200, $hdr, $body ]; +} + +sub set_filename ($$) { + my ($fn, $msg) = @_; + return to_filename($fn) if defined($fn); + + PublicInbox::Mbox::subject_fn($msg); +} + # called by Plack::Util::foreach or similar sub getline { my ($self) = @_; my $ctx = $self->{ctx} or return; - my $res; my $ibx = $ctx->{-inbox}; my $gz = $self->{gz}; + my $msgs = $self->{msgs}; do { - while (defined(my $smsg = shift @{$self->{msgs}})) { + # work on existing result set + while (defined(my $smsg = shift @$msgs)) { my $msg = eval { $ibx->msg_by_smsg($smsg) } or next; $msg = Email::Simple->new($msg); - $gz->write(PublicInbox::Mbox::msg_str($ctx, $msg)); + $gz->write(PublicInbox::Mbox::msg_str($ctx, $msg, + $smsg->mid)); + + # use subject of first message as subject + if (my $hdr = delete $self->{hdr}) { + my $fn = set_filename($self->{fn}, $msg); + push @$hdr, 'Content-Disposition', + "inline; filename=$fn.mbox.gz"; + return $hdr; + } my $bref = $self->{buf}; if (length($$bref) >= 8192) { my $ret = $$bref; # copy :< @@ -142,11 +245,11 @@ sub getline { # be fair to other clients on public-inbox-httpd: return ''; } - $res = $self->{cb}->($self->{opts}); - $self->{msgs} = $res->{msgs}; - $res = scalar @{$self->{msgs}}; - $self->{opts}->{offset} += $res; - } while ($res); + + # refill result set + $msgs = $self->{msgs} = $self->{cb}->($self->{opts}); + $self->{opts}->{offset} += scalar @$msgs; + } while (@$msgs); $gz->close; delete $self->{ctx}; ${delete $self->{buf}};