-# Copyright (C) 2015 all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2015-2018 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
# Streaming interface for formatting messages as an mboxrd.
# Used by the web interface
use strict;
use warnings;
use PublicInbox::MID qw/mid_clean mid_escape/;
-require Email::Simple;
+use PublicInbox::Hval qw/to_filename/;
+use Email::Simple;
+use Email::MIME::Encode;
-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, $msg)] ]
+sub subject_fn ($) {
+ 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
}
my $ibx = $ctx->{-inbox};
my $base = $ibx->base_url($ctx->{env});
- my $mid = mid_clean($header_obj->header('Message-ID'));
+ $mid = $ctx->{mid} unless defined $mid;
$mid = mid_escape($mid);
my @append = (
'Archived-At', "<$base$mid/>",
my ($ctx, $srch, $sfx) = @_;
eval { require IO::Compress::Gzip };
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) ];
+ my $prev = 0;
+ my $cb = sub {
+ my $msgs = $srch->get_thread($ctx->{mid}, $prev);
+ $prev = $msgs->[-1]->{num} if scalar(@$msgs);
+ $msgs;
+ };
+ PublicInbox::MboxGz->response($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 $@;
+ if ($query eq '') {
+ my $prev = 0;
+ my $msgs = [];
+ my $cb = sub {
+ $ctx->{-inbox}->mm->id_batch($prev, sub {
+ $msgs = $_[0];
+ });
+ $prev = $msgs->[-1] if @$msgs;
+ $msgs;
+ };
+ return PublicInbox::MboxGz->response($ctx, $cb, 'all');
+ }
+ my $opts = { offset => 0 };
+ my $srch = $ctx->{srch};
+ my $cb = sub { # called by MboxGz->getline
+ my $msgs = $srch->query($query, $opts);
+ $opts->{offset} += scalar @$msgs;
+ $msgs;
+ };
+ PublicInbox::MboxGz->response($ctx, $cb, 'results-'.$query);
}
sub need_gzip {
package PublicInbox::MboxGz;
use strict;
use warnings;
+use PublicInbox::Hval qw/to_filename/;
sub new {
my ($class, $ctx, $cb) = @_;
cb => $cb,
ctx => $ctx,
msgs => [],
- opts => { asc => 1, 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 ($$) {
+ my ($fn, $msg) = @_;
+ return to_filename($fn) if defined($fn);
+
+ PublicInbox::Mbox::subject_fn($msg);
+}
+
# called by Plack::Util::foreach or similar
sub getline {
my ($self) = @_;
my $ctx = $self->{ctx} or return;
- my $res;
my $ibx = $ctx->{-inbox};
my $gz = $self->{gz};
+ my $msgs = $self->{msgs};
do {
- while (defined(my $smsg = shift @{$self->{msgs}})) {
+ # work on existing result set
+ while (defined(my $smsg = shift @$msgs)) {
+ # id_batch may return integers
+ ref($smsg) or
+ $smsg = $ctx->{srch}->{over_ro}->get_art($smsg);
+
my $msg = eval { $ibx->msg_by_smsg($smsg) } or next;
$msg = Email::Simple->new($msg);
- $gz->write(PublicInbox::Mbox::msg_str($ctx, $msg));
+ $gz->write(PublicInbox::Mbox::msg_str($ctx, $msg,
+ $smsg->mid));
+
+ # use subject of first message as subject
+ if (my $hdr = delete $self->{hdr}) {
+ my $fn = set_filename($self->{fn}, $msg);
+ push @$hdr, 'Content-Disposition',
+ "inline; filename=$fn.mbox.gz";
+ return $hdr;
+ }
my $bref = $self->{buf};
if (length($$bref) >= 8192) {
my $ret = $$bref; # copy :<
# 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} = $self->{cb}->();
+ } while (@$msgs);
$gz->close;
+ # signal that we're done and can return undef next call:
delete $self->{ctx};
${delete $self->{buf}};
}