X-Git-Url: http://www.git.stargrave.org/?p=public-inbox.git;a=blobdiff_plain;f=lib%2FPublicInbox%2FNNTP.pm;h=5eb6112ceb2f932286bd263a11795b4c163590b1;hp=87ddf7a4af83312e41edc775e0794066d73e6ce3;hb=1c3e60b66e01df89afdf74990a849a5a7386f9c7;hpb=54c25a9cfb9a5eee44672b126eb75b9bb20aa957 diff --git a/lib/PublicInbox/NNTP.pm b/lib/PublicInbox/NNTP.pm index 87ddf7a4..5eb6112c 100644 --- a/lib/PublicInbox/NNTP.pm +++ b/lib/PublicInbox/NNTP.pm @@ -1,14 +1,15 @@ -# Copyright (C) 2015-2020 all contributors +# Copyright (C) all contributors # License: AGPL-3.0+ # # Each instance of this represents a NNTP client socket # fields: # nntpd: PublicInbox::NNTPD ref # article: per-session current article number -# ng: PublicInbox::Inbox ref +# ibx: PublicInbox::Inbox ref # long_cb: long_response private data package PublicInbox::NNTP; use strict; +use v5.10.1; use parent qw(PublicInbox::DS); use PublicInbox::MID qw(mid_escape $MID_EXTRACT); use PublicInbox::Eml; @@ -17,23 +18,23 @@ use PublicInbox::DS qw(now); use Digest::SHA qw(sha1_hex); use Time::Local qw(timegm timelocal); use PublicInbox::GitAsyncCat; +use PublicInbox::Address; + use constant { LINE_MAX => 512, # RFC 977 section 2.3 - r501 => '501 command syntax error', - r502 => '502 Command unavailable', - r221 => '221 Header follows', - r224 => '224 Overview information follows (multi-line)', - r225 => '225 Headers follow (multi-line)', - r430 => '430 No article with that message-id', + r501 => "501 command syntax error\r\n", + r502 => "502 Command unavailable\r\n", + r221 => "221 Header follows\r\n", + r225 => "225 Headers follow (multi-line)\r\n", + r430 => "430 No article with that message-id\r\n", }; -use PublicInbox::Syscall qw(EPOLLIN EPOLLONESHOT); use Errno qw(EAGAIN); my $ONE_MSGID = qr/\A$MID_EXTRACT\z/; my @OVERVIEW = qw(Subject From Date Message-ID References); my $OVERVIEW_FMT = join(":\r\n", @OVERVIEW, qw(Bytes Lines), '') . - "Xref:full\r\n"; + "Xref:full\r\n.\r\n"; my $LIST_HEADERS = join("\r\n", @OVERVIEW, - qw(:bytes :lines Xref To Cc)) . "\r\n"; + qw(:bytes :lines Xref To Cc)) . "\r\n.\r\n"; my $CAPABILITIES = <<""; 101 Capability list:\r VERSION 2\r @@ -44,32 +45,17 @@ HDR\r OVER\r COMPRESS DEFLATE\r -sub greet ($) { $_[0]->write($_[0]->{nntpd}->{greet}) }; - -sub new ($$$) { - my ($class, $sock, $nntpd) = @_; - my $self = bless { nntpd => $nntpd }, $class; - my $ev = EPOLLIN; - my $wbuf; - if ($sock->can('accept_SSL') && !$sock->accept_SSL) { - return CORE::close($sock) if $! != EAGAIN; - $ev = PublicInbox::TLS::epollbit(); - $wbuf = [ \&PublicInbox::DS::accept_tls_step, \&greet ]; - } - $self->SUPER::new($sock, $ev | EPOLLONESHOT); - if ($wbuf) { - $self->{wbuf} = $wbuf; - } else { - greet($self); - } - $self->update_idle_time; - $self; +sub do_greet ($) { $_[0]->write($_[0]->{nntpd}->{greet}) }; + +sub new { + my ($cls, $sock, $nntpd) = @_; + (bless { nntpd => $nntpd }, $cls)->greet($sock) } sub args_ok ($$) { my ($cb, $argc) = @_; my $tot = prototype $cb; - my ($nreq, undef) = split(';', $tot); + my ($nreq, undef) = split(/;/, $tot); $nreq = ($nreq =~ tr/$//) - 1; $tot = ($tot =~ tr/$//) - 1; ($argc <= $tot && $argc >= $nreq); @@ -80,20 +66,18 @@ sub process_line ($$) { my ($self, $l) = @_; my ($req, @args) = split(/[ \t]+/, $l); return 1 unless defined($req); # skip blank line - $req = $self->can('cmd_'.lc($req)); - return res($self, '500 command not recognized') unless $req; - return res($self, r501) unless args_ok($req, scalar @args); - + $req = $self->can('cmd_'.lc($req)) // + return $self->write(\"500 command not recognized\r\n"); + return $self->write(\r501) unless args_ok($req, scalar @args); my $res = eval { $req->($self, @args) }; my $err = $@; if ($err && $self->{sock}) { local $/ = "\n"; chomp($l); err($self, 'error from: %s (%s)', $l, $err); - $res = '503 program fault - command not performed'; + $res = \"503 program fault - command not performed\r\n"; } - return 0 unless defined $res; - res($self, $res); + defined($res) ? $self->write($res) : 0; } # The keyword argument is not used (rfc3977 5.2.2) @@ -104,62 +88,80 @@ sub cmd_capabilities ($;$) { $self->{nntpd}->{accept_tls}) { $res .= "STARTTLS\r\n"; } - $res .= '.'; + $res .= ".\r\n"; } sub cmd_mode ($$) { my ($self, $arg) = @_; - $arg = uc $arg; - return r501 unless $arg eq 'READER'; - '201 Posting prohibited'; + uc($arg) eq 'READER' ? \"201 Posting prohibited\r\n" : \r501; } -sub cmd_slave ($) { '202 slave status noted' } +sub cmd_slave ($) { \"202 slave status noted\r\n" } sub cmd_xgtitle ($;$) { my ($self, $wildmat) = @_; - more($self, '282 list of groups and descriptions follows'); + $self->msg_more("282 list of groups and descriptions follows\r\n"); list_newsgroups($self, $wildmat); - '.' } -sub list_overview_fmt ($) { - my ($self) = @_; - $self->msg_more($OVERVIEW_FMT); -} +sub list_overview_fmt ($) { $OVERVIEW_FMT } -sub list_headers ($;$) { - my ($self) = @_; - $self->msg_more($LIST_HEADERS); +sub list_headers ($;$) { $LIST_HEADERS } + +sub list_active_i { # "LIST ACTIVE" and also just "LIST" (no args) + my ($self, $groupnames) = @_; + my @window = splice(@$groupnames, 0, 100) or return 0; + my $ibx; + my $groups = $self->{nntpd}->{pi_cfg}->{-by_newsgroup}; + for my $ngname (@window) { + $ibx = $groups->{$ngname} and group_line($self, $ibx); + } + scalar(@$groupnames); # continue if there's more } -sub list_active ($;$) { +sub list_active ($;$) { # called by cmd_list my ($self, $wildmat) = @_; wildmat2re($wildmat); - foreach my $ng (@{$self->{nntpd}->{grouplist}}) { - $ng->{newsgroup} =~ $wildmat or next; - group_line($self, $ng); + long_response($self, \&list_active_i, [ + grep(/$wildmat/, @{$self->{nntpd}->{groupnames}}) ]); +} + +sub list_active_times_i { + my ($self, $groupnames) = @_; + my @window = splice(@$groupnames, 0, 100) or return 0; + my $groups = $self->{nntpd}->{pi_cfg}->{-by_newsgroup}; + for my $ngname (@window) { + my $ibx = $groups->{$ngname} or next; + my $c = eval { $ibx->uidvalidity } // time; + $self->msg_more("$ngname $c <$ibx->{-primary_address}>\r\n"); } + scalar(@$groupnames); # continue if there's more } -sub list_active_times ($;$) { +sub list_active_times ($;$) { # called by cmd_list my ($self, $wildmat) = @_; wildmat2re($wildmat); - foreach my $ng (@{$self->{nntpd}->{grouplist}}) { - $ng->{newsgroup} =~ $wildmat or next; - my $c = eval { $ng->mm->created_at } || time; - more($self, "$ng->{newsgroup} $c $ng->{-primary_address}"); + long_response($self, \&list_active_times_i, [ + grep(/$wildmat/, @{$self->{nntpd}->{groupnames}}) ]); +} + +sub list_newsgroups_i { + my ($self, $groupnames) = @_; + my @window = splice(@$groupnames, 0, 100) or return 0; + my $groups = $self->{nntpd}->{pi_cfg}->{-by_newsgroup}; + my $ibx; + for my $ngname (@window) { + $ibx = $groups->{$ngname} and + $self->msg_more("$ngname ".$ibx->description."\r\n"); } + scalar(@$groupnames); # continue if there's more } -sub list_newsgroups ($;$) { +sub list_newsgroups ($;$) { # called by cmd_list my ($self, $wildmat) = @_; wildmat2re($wildmat); - foreach my $ng (@{$self->{nntpd}->{grouplist}}) { - $ng->{newsgroup} =~ $wildmat or next; - my $d = $ng->description; - more($self, "$ng->{newsgroup} $d"); - } + long_response($self, \&list_newsgroups_i, [ + grep(/$wildmat/, @{$self->{nntpd}->{groupnames}}) ]); } # LIST SUBSCRIPTIONS, DISTRIB.PATS are not supported @@ -168,33 +170,32 @@ sub cmd_list ($;$$) { if (scalar @args) { my $arg = shift @args; $arg =~ tr/A-Z./a-z_/; + my $ret = $arg eq 'active'; $arg = "list_$arg"; $arg = $self->can($arg); return r501 unless $arg && args_ok($arg, scalar @args); - more($self, '215 information follows'); + $self->msg_more("215 information follows\r\n"); $arg->($self, @args); } else { - more($self, '215 list of newsgroups follows'); - foreach my $ng (@{$self->{nntpd}->{grouplist}}) { - group_line($self, $ng); - } + $self->msg_more("215 list of newsgroups follows\r\n"); + long_response($self, \&list_active_i, [ # copy array + @{$self->{nntpd}->{groupnames}} ]); } - '.' } sub listgroup_range_i { my ($self, $beg, $end) = @_; - my $r = $self->{ng}->mm->msg_range($beg, $end, 'num'); + my $r = $self->{ibx}->mm(1)->msg_range($beg, $end, 'num'); scalar(@$r) or return; - more($self, join("\r\n", map { $_->[0] } @$r)); + $self->msg_more(join("\r\n", @$r, '')); 1; } sub listgroup_all_i { my ($self, $num) = @_; - my $ary = $self->{ng}->mm->ids_after($num); + my $ary = $self->{ibx}->over(1)->ids_after($num); scalar(@$ary) or return; - more($self, join("\r\n", @$ary)); + $self->msg_more(join("\r\n", @$ary, '')); 1; } @@ -202,10 +203,10 @@ sub cmd_listgroup ($;$$) { my ($self, $group, $range) = @_; if (defined $group) { my $res = cmd_group($self, $group); - return $res if ($res !~ /\A211 /); - more($self, $res); + return $res if ref($res); # error if const strref + $self->msg_more($res); } - $self->{ng} or return '412 no newsgroup selected'; + $self->{ibx} or return \"412 no newsgroup selected\r\n"; if (defined $range) { my $r = get_range($self, $range); return $r unless ref $r; @@ -223,7 +224,7 @@ sub parse_time ($$;$) { $gmt = 1; } my ($YYYY, $MM, $DD); - if (bytes::length($date) == 8) { # RFC 3977 allows YYYYMMDD + if (length($date) == 8) { # RFC 3977 allows YYYYMMDD ($YYYY, $MM, $DD) = unpack('A4A2A2', $date); } else { # legacy clients send YYMMDD my $YY; @@ -242,9 +243,22 @@ sub parse_time ($$;$) { } sub group_line ($$) { - my ($self, $ng) = @_; - my ($min, $max) = $ng->mm->minmax; - more($self, "$ng->{newsgroup} $max $min n") if defined $min && defined $max; + my ($self, $ibx) = @_; + my ($min, $max) = $ibx->mm(1)->minmax; + $self->msg_more("$ibx->{newsgroup} $max $min n\r\n"); +} + +sub newgroups_i { + my ($self, $ts, $i, $groupnames) = @_; + my $end = $$i + 100; + my $groups = $self->{nntpd}->{pi_cfg}->{-by_newsgroup}; + while ($$i < $end) { + my $ngname = $groupnames->[$$i++] // return; + my $ibx = $groups->{$ngname} or next; # expired on reload + next unless (eval { $ibx->uidvalidity } // 0) > $ts; + group_line($self, $ibx); + } + 1; } sub cmd_newgroups ($$$;$$) { @@ -253,13 +267,9 @@ sub cmd_newgroups ($$$;$$) { return r501 if $@; # TODO dists - more($self, '231 list of new newsgroups follows'); - foreach my $ng (@{$self->{nntpd}->{grouplist}}) { - my $c = eval { $ng->mm->created_at } || 0; - next unless $c > $ts; - group_line($self, $ng); - } - '.' + $self->msg_more("231 list of new newsgroups follows\r\n"); + long_response($self, \&newgroups_i, $ts, \(my $i = 0), + $self->{nntpd}->{groupnames}); } sub wildmat2re (;$) { @@ -294,77 +304,71 @@ sub ngpat2re (;$) { } sub newnews_i { - my ($self, $overs, $ts, $prev) = @_; - my $over = $overs->[0]; - my $msgs = $over->query_ts($ts, $$prev); - if (scalar @$msgs) { - more($self, '<' . - join(">\r\n<", map { $_->{mid} } @$msgs ). - '>'); - $$prev = $msgs->[-1]->{num}; - } else { - shift @$overs; - if (@$overs) { # continue onto next newsgroup - $$prev = 0; - return 1; - } else { # break out of the long response. - return; + my ($self, $names, $ts, $prev) = @_; + my $ngname = $names->[0]; + if (my $ibx = $self->{nntpd}->{pi_cfg}->{-by_newsgroup}->{$ngname}) { + if (my $over = $ibx->over) { + my $msgs = $over->query_ts($ts, $$prev); + if (scalar @$msgs) { + $self->msg_more(join('', map { + "<$_->{mid}>\r\n"; + } @$msgs)); + $$prev = $msgs->[-1]->{num}; + return 1; # continue on current group + } } } + shift @$names; + if (@$names) { # continue onto next newsgroup + $$prev = 0; + 1; + } else { # all done, break out of the long_response + undef; + } } sub cmd_newnews ($$$$;$$) { my ($self, $newsgroups, $date, $time, $gmt, $dists) = @_; my $ts = eval { parse_time($date, $time, $gmt) }; return r501 if $@; - more($self, '230 list of new articles by message-id follows'); - my ($keep, $skip) = split('!', $newsgroups, 2); + $self->msg_more("230 list of new articles by message-id follows\r\n"); + my ($keep, $skip) = split(/!/, $newsgroups, 2); ngpat2re($keep); ngpat2re($skip); - my @overs; - foreach my $ng (@{$self->{nntpd}->{grouplist}}) { - $ng->{newsgroup} =~ $keep or next; - $ng->{newsgroup} =~ $skip and next; - my $over = $ng->over or next; - push @overs, $over; - }; - return '.' unless @overs; - + my @names = grep(!/$skip/, grep(/$keep/, + @{$self->{nntpd}->{groupnames}})); + return ".\r\n" unless scalar(@names); my $prev = 0; - long_response($self, \&newnews_i, \@overs, $ts, \$prev); + long_response($self, \&newnews_i, \@names, $ts, \$prev); } sub cmd_group ($$) { my ($self, $group) = @_; - my $no_such = '411 no such news group'; my $nntpd = $self->{nntpd}; - my $ng = $nntpd->{groups}->{$group} or return $no_such; + my $ibx = $nntpd->{pi_cfg}->{-by_newsgroup}->{$group} or + return \"411 no such news group\r\n"; $nntpd->idler_start; - $self->{ng} = $ng; - my ($min, $max) = $ng->mm->minmax; - $min ||= 0; - $max ||= 0; + $self->{ibx} = $ibx; + my ($min, $max) = $ibx->mm(1)->minmax; $self->{article} = $min; my $est_size = $max - $min; - "211 $est_size $min $max $group"; + "211 $est_size $min $max $group\r\n"; } sub article_adj ($$) { my ($self, $off) = @_; - my $ng = $self->{ng} or return '412 no newsgroup selected'; - - my $n = $self->{article}; - defined $n or return '420 no current article has been selected'; + my $ibx = $self->{ibx} // return \"412 no newsgroup selected\r\n"; + my $n = $self->{article} // + return \"420 no current article has been selected\r\n"; $n += $off; - my $mid = $ng->mm->mid_for($n); - unless ($mid) { + my $mid = $ibx->mm(1)->mid_for($n) // do { $n = $off > 0 ? 'next' : 'previous'; - return "421 no $n article in this group"; - } + return "421 no $n article in this group\r\n"; + }; $self->{article} = $n; - "223 $n <$mid> article retrieved - request text separately"; + "223 $n <$mid> article retrieved - request text separately\r\n"; } sub cmd_next ($) { article_adj($_[0], 1) } @@ -374,51 +378,61 @@ sub cmd_last ($) { article_adj($_[0], -1) } # the single-point-of-failure a single server provides. sub cmd_post ($) { my ($self) = @_; - my $ng = $self->{ng}; - $ng ? "440 mailto:$ng->{-primary_address} to post" - : '440 posting not allowed' + my $ibx = $self->{ibx}; + $ibx ? "440 mailto:$ibx->{-primary_address} to post\r\n" + : \"440 posting not allowed\r\n" } sub cmd_quit ($) { my ($self) = @_; - res($self, '205 closing connection - goodbye!'); + $self->write(\"205 closing connection - goodbye!\r\n"); $self->shutdn; undef; } -sub header_append ($$$) { - my ($hdr, $k, $v) = @_; - my @v = $hdr->header_raw($k); - foreach (@v) { - return if $v eq $_; +sub xref_by_tc ($$$) { + my ($xref, $pi_cfg, $smsg) = @_; + my $by_addr = $pi_cfg->{-by_addr}; + my $mid = $smsg->{mid}; + for my $f (qw(to cc)) { + my @ibxs = map { + $by_addr->{lc($_)} // () + } (PublicInbox::Address::emails($smsg->{$f} // '')); + for my $ibx (@ibxs) { + $xref->{$ibx->{newsgroup}} //= + $ibx->mm(1)->num_for($mid); + } } - $hdr->header_set($k, @v, $v); } -sub xref ($$$$) { - my ($self, $ng, $n, $mid) = @_; - my $ret = $self->{nntpd}->{servername} . " $ng->{newsgroup}:$n"; - - # num_for is pretty cheap and sometimes we'll lookup the existence - # of an article without getting even the OVER info. In other words, - # I'm not sure if its worth optimizing by scanning To:/Cc: and - # PublicInbox::ExtMsg on the PSGI end is just as expensive - foreach my $other (@{$self->{nntpd}->{grouplist}}) { - next if $ng eq $other; - my $num = eval { $other->mm->num_for($mid) } or next; - $ret .= " $other->{newsgroup}:$num"; +sub xref ($$$) { + my ($self, $cur_ibx, $smsg) = @_; + my $nntpd = $self->{nntpd}; + my $cur_ng = $cur_ibx->{newsgroup}; + my $xref; + if (my $ALL = $nntpd->{pi_cfg}->ALL) { + $xref = $ALL->nntp_xref_for($cur_ibx, $smsg); + xref_by_tc($xref, $nntpd->{pi_cfg}, $smsg); + } else { # slow path + $xref = { $cur_ng => $smsg->{num} }; + my $mid = $smsg->{mid}; + for my $ibx (values %{$nntpd->{pi_cfg}->{-by_newsgroup}}) { + $xref->{$ibx->{newsgroup}} //= + $ibx->mm(1)->num_for($mid); + } + } + my $ret = "$nntpd->{servername} $cur_ng:".delete($xref->{$cur_ng}); + for my $ng (sort keys %$xref) { + my $num = $xref->{$ng} // next; + $ret .= " $ng:$num"; } $ret; } -sub set_nntp_headers ($$$) { - my ($self, $hdr, $smsg) = @_; +sub set_nntp_headers ($$) { + my ($hdr, $smsg) = @_; my ($mid) = $smsg->{mid}; - # why? leafnode requires a Path: header for some inexplicable - # reason. We'll fake the shortest one possible. - $hdr->header_set('Path', 'y'); - # leafnode (and maybe other NNTP clients) have trouble dealing # with v2 messages which have multiple Message-IDs (either due # to our own content-based dedupe or buggy git-send-email versions). @@ -432,58 +446,59 @@ sub set_nntp_headers ($$$) { $hdr->header_set('X-Alt-Message-ID', @alt); } - # clobber some - my $ng = $self->{ng}; - my $xref = xref($self, $ng, $smsg->{num}, $mid); + # clobber some existing headers + my $ibx = $smsg->{-ibx}; + my $xref = xref($smsg->{nntp}, $ibx, $smsg); $hdr->header_set('Xref', $xref); - $xref =~ s/:[0-9]+//g; - $hdr->header_set('Newsgroups', (split(/ /, $xref, 2))[1]); - header_append($hdr, 'List-Post', "{-primary_address}>"); - if (my $url = $ng->base_url) { - $mid = mid_escape($mid); - header_append($hdr, 'Archived-At', "<$url$mid/>"); - header_append($hdr, 'List-Archive', "<$url>"); - } + + # RFC 5536 3.1.4 + my ($server_name, $newsgroups) = split(/ /, $xref, 2); + $newsgroups =~ s/:[0-9]+\b//g; # drop NNTP article numbers + $newsgroups =~ tr/ /,/; + $hdr->header_set('Newsgroups', $newsgroups); + + # *something* here is required for leafnode, try to follow + # RFC 5536 3.1.5... + $hdr->header_set('Path', $server_name . '!not-for-mail'); } -sub art_lookup ($$) { - my ($self, $art) = @_; - my $ng = $self->{ng}; - my ($n, $mid); +sub art_lookup ($$$) { + my ($self, $art, $code) = @_; + my ($ibx, $n); my $err; if (defined $art) { if ($art =~ /\A[0-9]+\z/) { - $err = '423 no such article number in this group'; + $err = \"423 no such article number in this group\r\n"; $n = int($art); - goto find_mid; + goto find_ibx; } elsif ($art =~ $ONE_MSGID) { - $mid = $1; - $err = r430; - $n = $ng->mm->num_for($mid) if $ng; - goto found if defined $n; - foreach my $g (values %{$self->{nntpd}->{groups}}) { - $n = $g->mm->num_for($mid); - if (defined $n) { - $ng = $g; - goto found; - } - } - return $err; + ($ibx, $n) = mid_lookup($self, $1); + goto found if $ibx; + return \r430; } else { - return r501; + return \r501; } } else { - $err = '420 no current article has been selected'; - $n = $self->{article}; - defined $n or return $err; -find_mid: - $ng or return '412 no newsgroup has been selected'; - $mid = $ng->mm->mid_for($n); - defined $mid or return $err; + $err = \"420 no current article has been selected\r\n"; + $n = $self->{article} // return $err; +find_ibx: + $ibx = $self->{ibx} or + return \"412 no newsgroup has been selected\r\n"; } found: - my $smsg = $ng->over->get_art($n) or return $err; - $smsg; + my $smsg = $ibx->over(1)->get_art($n) or return $err; + $smsg->{-ibx} = $ibx; + if ($code == 223) { # STAT + set_art($self, $n); + "223 $n <$smsg->{mid}> article retrieved - " . + "request text separately\r\n"; + } else { # HEAD | BODY | ARTICLE + $smsg->{nntp} = $self; + $smsg->{nntp_code} = $code; + set_art($self, $art); + # this dereferences to `undef' + ${ibx_async_cat($ibx, $smsg->{blob}, \&blob_cb, $smsg)}; + } } sub msg_body_write ($$) { @@ -501,9 +516,9 @@ sub set_art { $self->{article} = $art if defined $art && $art =~ /\A[0-9]+\z/; } -sub msg_hdr_write ($$$) { - my ($self, $eml, $smsg) = @_; - set_nntp_headers($self, $eml, $smsg); +sub msg_hdr_write ($$) { + my ($eml, $smsg) = @_; + set_nntp_headers($eml, $smsg); my $hdr = $eml->{hdr} // \(my $x = ''); # fixup old bug from import (pre-a0c07cba0e5d8b6a) @@ -513,17 +528,17 @@ sub msg_hdr_write ($$$) { # for leafnode compatibility, we need to ensure Message-ID headers # are only a single line. $$hdr =~ s/^(Message-ID:)[ \t]*\r\n[ \t]+([^\r]+)\r\n/$1 $2\r\n/igsm; - $self->msg_more($$hdr); + $smsg->{nntp}->msg_more($$hdr); } -sub blob_cb { # called by git->cat_async via git_async_cat +sub blob_cb { # called by git->cat_async via ibx_async_cat my ($bref, $oid, $type, $size, $smsg) = @_; my $self = $smsg->{nntp}; - my $code = $smsg->{nntp_code} // 220; + my $code = $smsg->{nntp_code}; if (!defined($oid)) { # it's possible to have TOCTOU if an admin runs # public-inbox-(edit|purge), just move onto the next message - warn "E: $smsg->{blob} missing in $self->{ng}->{inboxdir}\n"; + warn "E: $smsg->{blob} missing in $smsg->{-ibx}->{inboxdir}\n"; return $self->requeue; } elsif ($smsg->{blob} ne $oid) { $self->close; @@ -532,15 +547,15 @@ sub blob_cb { # called by git->cat_async via git_async_cat my $r = "$code $smsg->{num} <$smsg->{mid}> article retrieved - "; my $eml = PublicInbox::Eml->new($bref); if ($code == 220) { - more($self, $r .= 'head and body follow'); - msg_hdr_write($self, $eml, $smsg); + $self->msg_more($r .= "head and body follow\r\n"); + msg_hdr_write($eml, $smsg); $self->msg_more("\r\n"); msg_body_write($self, $bref); } elsif ($code == 221) { - more($self, $r .= 'head follows'); - msg_hdr_write($self, $eml, $smsg); + $self->msg_more($r .= "head follows\r\n"); + msg_hdr_write($eml, $smsg); } elsif ($code == 222) { - more($self, $r .= 'body follows'); + $self->msg_more($r .= "body follows\r\n"); msg_body_write($self, $bref); } else { $self->close; @@ -552,58 +567,38 @@ sub blob_cb { # called by git->cat_async via git_async_cat sub cmd_article ($;$) { my ($self, $art) = @_; - my $smsg = art_lookup($self, $art); - return $smsg unless ref $smsg; - set_art($self, $art); - $smsg->{nntp} = $self; - ${git_async_cat($self->{ng}->git, $smsg->{blob}, \&blob_cb, $smsg)}; + art_lookup($self, $art, 220); } sub cmd_head ($;$) { my ($self, $art) = @_; - my $smsg = art_lookup($self, $art); - return $smsg unless ref $smsg; - set_art($self, $art); - $smsg->{nntp} = $self; - $smsg->{nntp_code} = 221; - ${git_async_cat($self->{ng}->git, $smsg->{blob}, \&blob_cb, $smsg)}; + art_lookup($self, $art, 221); } sub cmd_body ($;$) { my ($self, $art) = @_; - my $smsg = art_lookup($self, $art); - return $smsg unless ref $smsg; - set_art($self, $art); - $smsg->{nntp} = $self; - $smsg->{nntp_code} = 222; - ${git_async_cat($self->{ng}->git, $smsg->{blob}, \&blob_cb, $smsg)}; + art_lookup($self, $art, 222); } sub cmd_stat ($;$) { my ($self, $art) = @_; - my $r = art_lookup($self, $art); - return $r unless ref $r; - my ($n, $mid) = @$r; - set_art($self, $art); - "223 $n <$mid> article retrieved - request text separately"; + art_lookup($self, $art, 223); # art may be msgid } -sub cmd_ihave ($) { '435 article not wanted - do not send it' } +sub cmd_ihave ($) { \"435 article not wanted - do not send it\r\n" } -sub cmd_date ($) { '111 '.strftime('%Y%m%d%H%M%S', gmtime(time)) } +sub cmd_date ($) { '111 '.strftime('%Y%m%d%H%M%S', gmtime(time))."\r\n" } -sub cmd_help ($) { - my ($self) = @_; - more($self, '100 help text follows'); - '.' -} +sub cmd_help ($) { \"100 help text follows\r\n.\r\n" } +# returns a ref on success sub get_range ($$) { my ($self, $range) = @_; - my $ng = $self->{ng} or return '412 no news group has been selected'; - defined $range or return '420 No article(s) selected'; + my $ibx = $self->{ibx} // + return "412 no news group has been selected\r\n"; + $range // return "420 No article(s) selected\r\n"; my ($beg, $end); - my ($min, $max) = $ng->mm->minmax; + my ($min, $max) = $ibx->mm(1)->minmax; if ($range =~ /\A([0-9]+)\z/) { $beg = $end = $1; } elsif ($range =~ /\A([0-9]+)-\z/) { @@ -615,8 +610,7 @@ sub get_range ($$) { } $beg = $min if ($beg < $min); $end = $max if ($end > $max); - return '420 No article(s) selected' if ($beg > $end); - [ \$beg, $end ]; + $beg > $end ? "420 No article(s) selected\r\n" : [ \$beg, $end ]; } sub long_step { @@ -635,22 +629,11 @@ sub long_step { out($self, " deferred[$fd] aborted - %0.6f", $elapsed); $self->close; } elsif ($more) { # $self->{wbuf}: - $self->update_idle_time; - - # COMPRESS users all share the same DEFLATE context. - # Flush it here to ensure clients don't see - # each other's data - $self->zflush; - - # no recursion, schedule another call ASAP, but only after - # all pending writes are done. autovivify wbuf: - my $new_size = push(@{$self->{wbuf}}, \&long_step); - - # wbuf may be populated by $cb, no need to rearm if so: - $self->requeue if $new_size == 1; + # control passed to ibx_async_cat if $more == \undef + $self->requeue_once if !ref($more); } else { # all done! delete $self->{long_cb}; - res($self, '.'); + $self->write(\".\r\n"); # TODO get rid of this my $elapsed = now() - $t0; my $fd = fileno($self->{sock}); out($self, " deferred[$fd] done - %0.6f", $elapsed); @@ -673,9 +656,9 @@ sub long_response ($$;@) { sub hdr_msgid_range_i { my ($self, $beg, $end) = @_; - my $r = $self->{ng}->mm->msg_range($beg, $end); + my $r = $self->{ibx}->mm(1)->msg_range($beg, $end); @$r or return; - more($self, join("\r\n", map { "$_->[0] <$_->[1]>" } @$r)); + $self->msg_more(join('', map { "$_->[0] <$_->[1]>\r\n" } @$r)); 1; } @@ -683,42 +666,68 @@ sub hdr_message_id ($$$) { # optimize XHDR Message-ID [range] for slrnpull. my ($self, $xhdr, $range) = @_; if (defined $range && $range =~ $ONE_MSGID) { - my ($ng, $n) = mid_lookup($self, $1); + my ($ibx, $n) = mid_lookup($self, $1); return r430 unless $n; - hdr_mid_response($self, $xhdr, $ng, $n, $range, $range); + hdr_mid_response($self, $xhdr, $ibx, $n, $range, $range); } else { # numeric range $range = $self->{article} unless defined $range; my $r = get_range($self, $range); return $r unless ref $r; - more($self, $xhdr ? r221 : r225); + $self->msg_more($xhdr ? r221 : r225); long_response($self, \&hdr_msgid_range_i, @$r); } } sub mid_lookup ($$) { my ($self, $mid) = @_; - my $self_ng = $self->{ng}; - if ($self_ng) { - my $n = $self_ng->mm->num_for($mid); - return ($self_ng, $n) if defined $n; + my $cur_ibx = $self->{ibx}; + if ($cur_ibx) { + my $n = $cur_ibx->mm(1)->num_for($mid); + return ($cur_ibx, $n) if defined $n; } - foreach my $ng (values %{$self->{nntpd}->{groups}}) { - next if defined $self_ng && $ng eq $self_ng; - my $n = $ng->mm->num_for($mid); - return ($ng, $n) if defined $n; + my $pi_cfg = $self->{nntpd}->{pi_cfg}; + if (my $ALL = $pi_cfg->ALL) { + my ($id, $prev); + while (my $smsg = $ALL->over->next_by_mid($mid, \$id, \$prev)) { + my $xr3 = $ALL->over->get_xref3($smsg->{num}); + if (my @x = grep(/:$smsg->{blob}\z/, @$xr3)) { + my ($ngname, $xnum) = split(/:/, $x[0]); + my $ibx = $pi_cfg->{-by_newsgroup}->{$ngname}; + return ($ibx, $xnum) if $ibx; + # fall through to trying all xref3s + } else { + warn < ($smsg->{blob}) in $ALL->{topdir}, -extindex bug? +EOF + } + # try all xref3s + for my $x (@$xr3) { + my ($ngname, $xnum) = split(/:/, $x); + my $ibx = $pi_cfg->{-by_newsgroup}->{$ngname}; + return ($ibx, $xnum) if $ibx; + warn "W: `$ngname' does not exist for #$xnum\n"; + } + } + # no warning here, $mid is just invalid + } else { # slow path for non-ALL users + for my $ibx (values %{$pi_cfg->{-by_newsgroup}}) { + next if defined $cur_ibx && $ibx eq $cur_ibx; + my $n = $ibx->mm(1)->num_for($mid); + return ($ibx, $n) if defined $n; + } } (undef, undef); } sub xref_range_i { my ($self, $beg, $end) = @_; - my $ng = $self->{ng}; - my $r = $ng->mm->msg_range($beg, $end); - @$r or return; - more($self, join("\r\n", map { - my $num = $_->[0]; - "$num ".xref($self, $ng, $num, $_->[1]); - } @$r)); + my $ibx = $self->{ibx}; + my $msgs = $ibx->over(1)->query_xover($$beg, $end); + scalar(@$msgs) or return; + $$beg = $msgs->[-1]->{num} + 1; + $self->msg_more(join('', map { + "$_->{num} ".xref($self, $ibx, $_) . "\r\n"; + } @$msgs)); 1; } @@ -727,30 +736,30 @@ sub hdr_xref ($$$) { # optimize XHDR Xref [range] for rtin if (defined $range && $range =~ $ONE_MSGID) { my $mid = $1; - my ($ng, $n) = mid_lookup($self, $mid); + my ($ibx, $n) = mid_lookup($self, $mid); return r430 unless $n; - hdr_mid_response($self, $xhdr, $ng, $n, $range, - xref($self, $ng, $n, $mid)); + my $smsg = $ibx->over(1)->get_art($n) or return; + hdr_mid_response($self, $xhdr, $ibx, $n, $range, + xref($self, $ibx, $smsg)); } else { # numeric range $range = $self->{article} unless defined $range; my $r = get_range($self, $range); return $r unless ref $r; - more($self, $xhdr ? r221 : r225); + $self->msg_more($xhdr ? r221 : r225); long_response($self, \&xref_range_i, @$r); } } sub over_header_for { - my ($over, $num, $field) = @_; - my $smsg = $over->get_art($num) or return; + my ($ibx, $num, $field) = @_; + my $smsg = $ibx->over(1)->get_art($num) or return; return PublicInbox::Smsg::date($smsg) if $field eq 'date'; $smsg->{$field}; } sub smsg_range_i { my ($self, $beg, $end, $field) = @_; - my $over = $self->{ng}->over; - my $msgs = $over->query_xover($$beg, $end); + my $msgs = $self->{ibx}->over(1)->query_xover($$beg, $end); scalar(@$msgs) or return; my $tmp = ''; @@ -772,15 +781,15 @@ sub smsg_range_i { sub hdr_smsg ($$$$) { my ($self, $xhdr, $field, $range) = @_; if (defined $range && $range =~ $ONE_MSGID) { - my ($ng, $n) = mid_lookup($self, $1); + my ($ibx, $n) = mid_lookup($self, $1); return r430 unless defined $n; - my $v = over_header_for($ng->over, $n, $field); - hdr_mid_response($self, $xhdr, $ng, $n, $range, $v); + my $v = over_header_for($ibx, $n, $field); + hdr_mid_response($self, $xhdr, $ibx, $n, $range, $v); } else { # numeric range $range = $self->{article} unless defined $range; my $r = get_range($self, $range); return $r unless ref $r; - more($self, $xhdr ? r221 : r225); + $self->msg_more($xhdr ? r221 : r225); long_response($self, \&smsg_range_i, @$r, $field); } } @@ -798,7 +807,7 @@ sub do_hdr ($$$;$) { } elsif ($sub =~ /\A:(bytes|lines)\z/) { hdr_smsg($self, $xhdr, $1, $range); } else { - $xhdr ? (r221 . "\r\n.") : "503 HDR not permitted on $header"; + $xhdr ? (r221.".\r\n") : "503 HDR not permitted on $header\r\n"; } } @@ -815,57 +824,50 @@ sub cmd_xhdr ($$;$) { } sub hdr_mid_prefix ($$$$$) { - my ($self, $xhdr, $ng, $n, $mid) = @_; + my ($self, $xhdr, $ibx, $n, $mid) = @_; return $mid if $xhdr; # HDR for RFC 3977 users - if (my $self_ng = $self->{ng}) { - ($self_ng eq $ng) ? $n : '0'; + if (my $cur_ibx = $self->{ibx}) { + ($cur_ibx eq $ibx) ? $n : '0'; } else { '0'; } } sub hdr_mid_response ($$$$$$) { - my ($self, $xhdr, $ng, $n, $mid, $v) = @_; - my $res = ''; - if ($xhdr) { - $res .= r221 . "\r\n"; - $res .= "$mid $v\r\n"; - } else { - $res .= r225 . "\r\n"; - my $pfx = hdr_mid_prefix($self, $xhdr, $ng, $n, $mid); - $res .= "$pfx $v\r\n"; - } - res($self, $res .= '.'); + my ($self, $xhdr, $ibx, $n, $mid, $v) = @_; + $self->write(($xhdr ? r221.$mid : + r225.hdr_mid_prefix($self, $xhdr, $ibx, $n, $mid)) . + " $v\r\n.\r\n"); undef; } sub xrover_i { my ($self, $beg, $end) = @_; - my $h = over_header_for($self->{ng}->over, $$beg, 'references'); - more($self, "$$beg $h") if defined($h); + my $h = over_header_for($self->{ibx}, $$beg, 'references'); + $self->msg_more("$$beg $h\r\n") if defined($h); $$beg++ < $end; } sub cmd_xrover ($;$) { my ($self, $range) = @_; - my $ng = $self->{ng} or return '412 no newsgroup selected'; + my $ibx = $self->{ibx} or return \"412 no newsgroup selected\r\n"; (defined $range && $range =~ /[<>]/) and - return '420 No article(s) selected'; # no message IDs + return \"420 No article(s) selected\r\n"; # no message IDs $range = $self->{article} unless defined $range; my $r = get_range($self, $range); return $r unless ref $r; - more($self, '224 Overview information follows'); + $self->msg_more("224 Overview information follows\r\n"); long_response($self, \&xrover_i, @$r); } -sub over_line ($$$$) { - my ($self, $ng, $num, $smsg) = @_; +sub over_line ($$$) { + my ($self, $ibx, $smsg) = @_; # n.b. field access and procedural calls can be # 10%-15% faster than OO method calls: - my $s = join("\t", $num, + my $s = join("\t", $smsg->{num}, $smsg->{subject}, $smsg->{from}, PublicInbox::Smsg::date($smsg), @@ -873,24 +875,29 @@ sub over_line ($$$$) { $smsg->{references}, $smsg->{bytes}, $smsg->{lines}, - "Xref: " . xref($self, $ng, $num, $smsg->{mid})); + "Xref: " . xref($self, $ibx, $smsg)); utf8::encode($s); - $s + $s .= "\r\n"; } sub cmd_over ($;$) { my ($self, $range) = @_; if ($range && $range =~ $ONE_MSGID) { - my ($ng, $n) = mid_lookup($self, $1); + my ($ibx, $n) = mid_lookup($self, $1); defined $n or return r430; - my $smsg = $ng->over->get_art($n) or return r430; - more($self, '224 Overview information follows (multi-line)'); + my $smsg = $ibx->over(1)->get_art($n) or return r430; + $self->msg_more( + "224 Overview information follows (multi-line)\r\n"); # Only set article number column if it's the current group - my $self_ng = $self->{ng}; - $n = 0 if (!$self_ng || $self_ng ne $ng); - more($self, over_line($self, $ng, $n, $smsg)); - '.'; + # (RFC 3977 8.3.2) + my $cur_ibx = $self->{ibx}; + if (!$cur_ibx || $cur_ibx ne $ibx) { + # set {-orig_num} for nntp_xref_for + $smsg->{-orig_num} = $smsg->{num}; + $smsg->{num} = 0; + } + over_line($self, $ibx, $smsg).".\r\n"; } else { cmd_xover($self, $range); } @@ -898,13 +905,13 @@ sub cmd_over ($;$) { sub xover_i { my ($self, $beg, $end) = @_; - my $ng = $self->{ng}; - my $msgs = $ng->over->query_xover($$beg, $end); + my $ibx = $self->{ibx}; + my $msgs = $ibx->over(1)->query_xover($$beg, $end); my $nr = scalar @$msgs or return; # OVERVIEW.FMT - more($self, join("\r\n", map { - over_line($self, $ng, $_->{num}, $_); + $self->msg_more(join('', map { + over_line($self, $ibx, $_); } @$msgs)); $$beg = $msgs->[-1]->{num} + 1; } @@ -915,7 +922,8 @@ sub cmd_xover ($;$) { my $r = get_range($self, $range); return $r unless ref $r; my ($beg, $end) = @$r; - more($self, "224 Overview information follows for $$beg to $end"); + $self->msg_more( + "224 Overview information follows for $$beg to $end\r\n"); long_response($self, \&xover_i, @$r); } @@ -927,8 +935,8 @@ sub cmd_starttls ($) { # RFC 4642 2.2.1 return r502 if ($sock->can('accept_SSL') || $self->compressed); my $opt = $self->{nntpd}->{accept_tls} or - return '580 can not initiate TLS negotiation'; - res($self, '382 Continue with TLS negotiation'); + return \"580 can not initiate TLS negotiation\r\n"; + $self->write(\"382 Continue with TLS negotiation\r\n"); $self->{sock} = IO::Socket::SSL->start_SSL($sock, %$opt); $self->requeue if PublicInbox::DS::accept_tls_step($self); undef; @@ -937,38 +945,40 @@ sub cmd_starttls ($) { # RFC 8054 sub cmd_compress ($$) { my ($self, $alg) = @_; - return '503 Only DEFLATE is supported' if uc($alg) ne 'DEFLATE'; + return "503 Only DEFLATE is supported\r\n" if uc($alg) ne 'DEFLATE'; return r502 if $self->compressed; PublicInbox::NNTPdeflate->enable($self); $self->requeue; undef } -sub zflush {} # overridden by NNTPdeflate - sub cmd_xpath ($$) { my ($self, $mid) = @_; return r501 unless $mid =~ $ONE_MSGID; $mid = $1; my @paths; - foreach my $ng (values %{$self->{nntpd}->{groups}}) { - my $n = $ng->mm->num_for($mid); - push @paths, "$ng->{newsgroup}/$n" if defined $n; + my $pi_cfg = $self->{nntpd}->{pi_cfg}; + my $groups = $pi_cfg->{-by_newsgroup}; + if (my $ALL = $pi_cfg->ALL) { + my ($id, $prev, %seen); + while (my $smsg = $ALL->over->next_by_mid($mid, \$id, \$prev)) { + my $xr3 = $ALL->over->get_xref3($smsg->{num}); + for my $x (@$xr3) { + my ($ngname, $n) = split(/:/, $x); + $x = "$ngname/$n"; + if ($groups->{$ngname} && !$seen{$x}++) { + push(@paths, $x); + } + } + } + } else { # slow path, no point in using long_response + for my $ibx (values %$groups) { + my $n = $ibx->mm(1)->num_for($mid) // next; + push @paths, "$ibx->{newsgroup}/$n"; + } } - return '430 no such article on server' unless @paths; - '223 '.join(' ', @paths); -} - -sub res ($$) { do_write($_[0], $_[1] . "\r\n") } - -sub more ($$) { $_[0]->msg_more($_[1] . "\r\n") } - -sub do_write ($$) { - my $self = $_[0]; - my $done = $self->write(\($_[1])); - return 0 unless $self->{sock}; - - $done; + return \"430 no such article on server\r\n" unless @paths; + '223 '.join(' ', sort(@paths))."\r\n"; } sub err ($$;@) { @@ -987,7 +997,6 @@ sub event_step { return unless $self->flush_write && $self->{sock} && !$self->{long_cb}; - $self->update_idle_time; # only read more requests if we've drained the write buffer, # otherwise we can be buffering infinitely w/o backpressure @@ -1009,17 +1018,15 @@ sub event_step { out($self, "[$fd] %s - %0.6f$pending", $line, now() - $t0); return $self->close if $r < 0; $self->rbuf_idle($rbuf); - $self->update_idle_time; # maybe there's more pipelined data, or we'll have # to register it for socket-readiness notifications $self->requeue unless $pending; } -# for graceful shutdown in PublicInbox::Daemon: -sub busy { - my ($self, $now) = @_; - ($self->{rbuf} || $self->{wbuf} || $self->not_idle_long($now)); +sub busy { # for graceful shutdown in PublicInbox::Daemon: + my ($self) = @_; + defined($self->{rbuf}) || defined($self->{wbuf}) } 1;