X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FMbox.pm;h=04c86cc196515e345d36491b200a1d29f3e4bd17;hb=15eb65ffd59c7cf5cce67c6b7621f63194fd1083;hp=1c792c893a3ea6ee4099f9ff9fea2193a1a44f81;hpb=1e2f280d973dd0a744d134368fa9fbfbb6aab161;p=public-inbox.git diff --git a/lib/PublicInbox/Mbox.pm b/lib/PublicInbox/Mbox.pm index 1c792c89..04c86cc1 100644 --- a/lib/PublicInbox/Mbox.pm +++ b/lib/PublicInbox/Mbox.pm @@ -1,30 +1,53 @@ -# 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/; -use Plack::Util; -require Email::Simple; +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 emit1 { - my $simple = Email::Simple->new(pop); - my $ctx = pop; - sub { - my ($response) = @_; - # single message should be easily renderable in browsers - my $fh = $response->([200, ['Content-Type'=>'text/plain']]); - emit_msg($ctx, $fh, $simple); - $fh->close; + my ($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, $msg) ] ] } -sub emit_msg { - my ($ctx, $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 @@ -33,16 +56,17 @@ sub emit_msg { $header_obj->header_set($d); } my $ibx = $ctx->{-inbox}; - my $base = $ibx->base_url($ctx->{cgi}); + my $base = $ibx->base_url($ctx->{env}); my $mid = mid_clean($header_obj->header('Message-ID')); - $mid = uri_escape_utf8($mid); + $mid = mid_escape($mid); my @append = ( 'Archived-At', "<$base$mid/>", 'List-Archive', "<$base>", 'List-Post', "{-primary_address}>", ); - my $append = ''; 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]; @@ -53,54 +77,46 @@ sub emit_msg { last; } } - $append .= "$k: $v$crlf" if defined $v; + $buf .= "$k: $v$crlf" if defined $v; } - my $buf = $header_obj->as_string; - unless ($buf =~ /\AFrom /) { - $fh->write("From mboxrd\@z Thu Jan 1 00:00:00 1970\n"); - } - $append .= $crlf; - $fh->write($buf .= $append); - - $buf = $simple->body; - $simple->body_set(''); + $buf .= $crlf; # mboxrd quoting style # ref: http://www.qmail.org/man/man5/mbox.html - $buf =~ s/^(>*From )/>$1/gm; - - $fh->write($buf .= "\n"); + my $body = $simple->body; + $body =~ s/^(>*From )/>$1/gm; + $buf .= $body; + $buf .= "\n"; } -sub noop {} - sub thread_mbox { my ($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, $cb) ]; + PublicInbox::MboxGz->response($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 { @@ -120,54 +136,84 @@ EOF package PublicInbox::MboxGz; use strict; use warnings; -use PublicInbox::MID qw(mid2path); +use PublicInbox::Hval qw/to_filename/; sub new { my ($class, $ctx, $cb) = @_; - my $buf; + my $buf = ''; bless { buf => \$buf, gz => IO::Compress::Gzip->new(\$buf, Time => 0), cb => $cb, ctx => $ctx, msgs => [], - opts => { asc => 1, offset => 0 }, + opts => { offset => 0 }, }, $class; } -sub _flush_buf { - my ($self) = @_; - my $ret = $self->{buf}; - $ret = $$ret; - ${$self->{buf}} = undef; - $ret; +sub response { + my ($class, $ctx, $cb, $fn) = @_; + my $body = $class->new($ctx, $cb); + # http://www.iana.org/assignments/media-types/application/gzip + $body->{hdr} = [ 'Content-Type', 'application/gzip' ]; + $body->{fn} = $fn; + # fill in Content-Disposition filename + my $hdr = $body->getline; + if ($body->{hdr}) { + return [ 404, ['Content-Type','text/plain'], + [ "No results found\n" ] ]; + } + [ 200, $hdr, $body ]; +} + +sub set_filename ($$) { + 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 $ctx = $self->{ctx}; - my $git = $ctx->{git}; + my $ibx = $ctx->{-inbox}; + my $gz = $self->{gz}; do { + # work on existing result set while (defined(my $smsg = shift @{$self->{msgs}})) { - my $msg = eval { - my $p = 'HEAD:'.mid2path($smsg->mid); - Email::Simple->new($git->cat_file($p)); - }; - $msg or next; - - PublicInbox::Mbox::emit_msg($ctx, $self->{gz}, $msg); - my $ret = _flush_buf($self); - return $ret if $ret; + 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); - $self->{gz}->close; - _flush_buf($self); + $gz->close; + delete $self->{ctx}; + ${delete $self->{buf}}; } sub close {} # noop