diff options
Diffstat (limited to 'lib/PublicInbox/LeiXSearch.pm')
-rw-r--r-- | lib/PublicInbox/LeiXSearch.pm | 324 |
1 files changed, 148 insertions, 176 deletions
diff --git a/lib/PublicInbox/LeiXSearch.pm b/lib/PublicInbox/LeiXSearch.pm index 6f877019..43dedd10 100644 --- a/lib/PublicInbox/LeiXSearch.pm +++ b/lib/PublicInbox/LeiXSearch.pm @@ -12,15 +12,17 @@ use PublicInbox::DS qw(now); use File::Temp 0.19 (); # 0.19 for ->newdir use File::Spec (); use PublicInbox::Search qw(xap_terms); -use PublicInbox::Spawn qw(popen_rd spawn which); -use PublicInbox::MID qw(mids); +use PublicInbox::Spawn qw(popen_rd popen_wr which); +use PublicInbox::MID qw(mids mid_escape); use PublicInbox::Smsg; -use PublicInbox::AutoReap; use PublicInbox::Eml; use PublicInbox::LEI; use Fcntl qw(SEEK_SET F_SETFL O_APPEND O_RDWR); use PublicInbox::ContentHash qw(git_sha); use POSIX qw(strftime); +use autodie qw(close open read seek truncate); +use PublicInbox::Syscall qw($F_SETPIPE_SZ); +use PublicInbox::OnDestroy; sub new { my ($class) = @_; @@ -103,13 +105,6 @@ sub smsg_for { $smsg; } -sub recent { - my ($self, $qstr, $opt) = @_; - $opt //= {}; - $opt->{relevance} //= -2; - $self->mset($qstr //= 'z:1..', $opt); -} - sub over {} sub _check_mset_limit ($$$) { @@ -128,26 +123,16 @@ sub _mset_more ($$) { $size >= $mo->{limit} && (($mo->{offset} += $size) < $mo->{total}); } -# $startq will EOF when do_augment is done augmenting and allow +# $startq will see `q' in do_post_augment -> start_mua if spawning MUA. +# Otherwise $startq will EOF when do_augment is done augmenting and allow # query_combined_mset and query_thread_mset to proceed. sub wait_startq ($) { my ($lei) = @_; - my $startq = delete $lei->{startq} or return; - while (1) { - my $n = sysread($startq, my $do_augment_done, 1); - if (defined $n) { - return if $n == 0; # no MUA - if ($do_augment_done eq 'q') { - $lei->{opt}->{quiet} = 1; - delete $lei->{opt}->{verbose}; - delete $lei->{-progress}; - } else { - die "BUG: do_augment_done=`$do_augment_done'"; - } - return; - } - die "wait_startq: $!" unless $!{EINTR}; - } + read(delete($lei->{startq}) // return, my $buf, 1) or return; # EOF + die "BUG: wrote `$buf' to au_done" if $buf ne 'q'; + $lei->{opt}->{quiet} = 1; + delete $lei->{opt}->{verbose}; + delete $lei->{-progress}; } sub mset_progress { @@ -162,71 +147,62 @@ sub mset_progress { } } -sub l2m_progress { - my ($lei, $nr_write, $nr_seen) = @_; - $lei->{-nr_write} += $nr_write; - $lei->{-nr_seen} += $nr_seen; -} - sub query_one_mset { # for --threads and l2m w/o sort my ($self, $ibxish) = @_; - local $0 = "$0 query_one_mset"; my $lei = $self->{lei}; my ($srch, $over) = ($ibxish->search, $ibxish->over); my $dir = $ibxish->{inboxdir} // $ibxish->{topdir}; return warn("$dir not indexed by Xapian\n") unless ($srch && $over); bless $srch, 'PublicInbox::LeiSearch'; # for ->qparse_new my $mo = { %{$lei->{mset_opt}} }; # copy + local $0 = "$0 1 $mo->{qstr}"; my $mset; my $each_smsg = $lei->{ovv}->ovv_each_smsg_cb($lei); my $can_kw = !!$ibxish->can('msg_keywords'); my $threads = $lei->{opt}->{threads} // 0; my $fl = $threads > 1 ? 1 : undef; + my $mid = $lei->{opt}->{'thread-id'}; + $mo->{threadid} = $over->mid2tid($mid) if defined $mid; my $lss = $lei->{lss}; - my $maxk = "external.$dir.maxuid"; - my $stop_at = $lss ? $lss->{-cfg}->{$maxk} : undef; - if (defined $stop_at) { - ref($stop_at) and - return warn("$maxk=$stop_at has multiple values\n"); - ($stop_at =~ /[^0-9]/) and - return warn("$maxk=$stop_at not numeric\n"); - } + my $maxk = "external.$dir.maxuid"; # max of previous, so our min + my $min = $lss ? ($lss->{-cfg}->{$maxk} // 0) : 0; + ref($min) and return warn("$maxk=$min has multiple values\n"); + ($min =~ /[^0-9]/) and return warn("$maxk=$min not numeric\n"); my $first_ids; do { - $mset = $srch->mset($mo->{qstr}, $mo); + $mset = eval { $srch->mset($mo->{qstr}, $mo) }; + return $lei->child_error(22 << 8, "E: $@") if $@; # 22 from curl mset_progress($lei, $dir, $mo->{offset} + $mset->size, $mset->get_matches_estimated); wait_startq($lei); # wait for keyword updates my $ids = $srch->mset_to_artnums($mset, $mo); - @$ids = grep { $_ > $stop_at } @$ids if defined($stop_at); my $i = 0; if ($threads) { # copy $ids if $lss since over->expand_thread # shifts @{$ctx->{ids}} $first_ids = [ @$ids ] if $lss; - my $ctx = { ids => $ids }; - my %n2item = map { ($ids->[$i++], $_) } $mset->items; - while ($over->expand_thread($ctx)) { - for my $n (@{$ctx->{xids}}) { + my $ctx = { ids => $ids, min => $min }; + my %n2item = map { $ids->[$i++] => $_ } $mset->items; + while ($over->expand_thread($ctx)) { # fills {xids} + for my $n (@{delete $ctx->{xids}}) { my $smsg = $over->get_art($n) or next; - my $mitem = delete $n2item{$n}; + my $mi = delete $n2item{$n}; next if $smsg->{bytes} == 0; - if ($mitem && $can_kw) { - mitem_kw($srch, $smsg, $mitem, - $fl); - } elsif ($mitem && $fl) { + if ($mi && $can_kw) { + mitem_kw($srch, $smsg, $mi, $fl) + } elsif ($mi && $fl) { # call ->xsmsg_vmd, later $smsg->{lei_q_tt_flagged} = 1; } - $each_smsg->($smsg, $mitem); + $each_smsg->($smsg, $mi); } - @{$ctx->{xids}} = (); } } else { $first_ids = $ids; - my @items = $mset->items; + my @items = $mset->items; # parallel with @$ids for my $n (@$ids) { my $mitem = $items[$i++]; + next if $n <= $min; my $smsg = $over->get_art($n) or next; next if $smsg->{bytes} == 0; mitem_kw($srch, $smsg, $mitem, $fl) if $can_kw; @@ -236,7 +212,6 @@ sub query_one_mset { # for --threads and l2m w/o sort } while (_mset_more($mset, $mo)); _check_mset_limit($lei, $dir, $mset); if ($lss && scalar(@$first_ids)) { - undef $stop_at; my $max = $first_ids->[0]; $lss->cfg_set($maxk, $max); undef $lss; @@ -247,16 +222,17 @@ sub query_one_mset { # for --threads and l2m w/o sort sub query_combined_mset { # non-parallel for non-"--threads" users my ($self) = @_; - local $0 = "$0 query_combined_mset"; my $lei = $self->{lei}; my $mo = { %{$lei->{mset_opt}} }; + local $0 = "$0 C $mo->{qstr}"; my $mset; for my $loc (locals($self)) { attach_external($self, $loc); } my $each_smsg = $lei->{ovv}->ovv_each_smsg_cb($lei); do { - $mset = $self->mset($mo->{qstr}, $mo); + $mset = eval { $self->mset($mo->{qstr}, $mo) }; + return $lei->child_error(22 << 8, "E: $@") if $@; # 22 from curl mset_progress($lei, 'xsearch', $mo->{offset} + $mset->size, $mset->get_matches_estimated); wait_startq($lei); # wait for keyword updates @@ -325,8 +301,9 @@ sub fudge_qstr_time ($$$) { $rft = $diff; } $lr -= ($rft || (48 * 60 * 60)); + require PublicInbox::Admin; $lei->qerr("# $uri limiting to ". - strftime('%Y-%m-%d %k:%M %z', localtime($lr)). ' and newer'); + PublicInbox::Admin::fmt_localtime($lr).' and newer'); # this should really be rt: (received-time), but no stable # public-inbox releases support it, yet. my $dt = 'dt:'.strftime('%Y%m%d%H%M%S', gmtime($lr)).'..'; @@ -339,67 +316,69 @@ sub fudge_qstr_time ($$$) { sub query_remote_mboxrd { my ($self, $uris) = @_; - local $0 = "$0 query_remote_mboxrd"; local $SIG{TERM} = sub { exit(0) }; # for DESTROY (File::Temp, $reap) my $lei = $self->{lei}; my $opt = $lei->{opt}; - chomp(my $qstr = $lei->{mset_opt}->{qstr}); - $qstr =~ s/[ \n\t]+/ /sg; # make URLs less ugly + my $qstr = $lei->{mset_opt}->{qstr}; + local $0 = "$0 R $qstr"; my @qform = (x => 'm'); push(@qform, t => 1) if $opt->{threads}; - my $verbose = $opt->{verbose}; - my $reap_tail; - my $cerr = File::Temp->new(TEMPLATE => 'curl.err-XXXX', TMPDIR => 1); - fcntl($cerr, F_SETFL, O_APPEND|O_RDWR) or warn "set O_APPEND: $!"; + open my $cerr, '+>', undef; my $rdr = { 2 => $cerr }; - if ($verbose) { - # spawn a process to force line-buffering, otherwise curl + my @lbf_tee; + if ($opt->{verbose}) { + # spawn a line-buffered tee(1) script, otherwise curl # will write 1 character at-a-time and parallel outputs # mmmaaayyy llloookkk llliiikkkeee ttthhhiiisss - my $o = { 1 => $lei->{2}, 2 => $lei->{2} }; - my $pid = spawn(['tail', '-f', $cerr->filename], undef, $o); - $reap_tail = PublicInbox::AutoReap->new($pid); + # (n.b. POSIX tee(1) cannot do any buffering) + my $o = { 1 => $cerr, 2 => $lei->{2} }; + delete $rdr->{2}; + @lbf_tee = ([ $^X, qw(-w -p -e), <<'' ], undef, $o); +BEGIN { $| = 1; use IO::Handle; STDERR->autoflush(1); } +print STDERR $_; + } my $curl = PublicInbox::LeiCurl->new($lei, $self->{curl}) or return; push @$curl, '-s', '-d', ''; my $each_smsg = $lei->{ovv}->ovv_each_smsg_cb($lei); $self->{import_sto} = $lei->{sto} if $lei->{opt}->{'import-remote'}; + if (defined(my $mid = $opt->{'thread-id'})) { + $mid = mid_escape($mid); + for my $uri (@$uris) { + $uri->path($uri->path.$mid.'/'); + } + } for my $uri (@$uris) { $lei->{-current_url} = $uri->as_string; - $lei->{-nr_remote_eml} = 0; my $start = time; my ($q, $key) = fudge_qstr_time($lei, $uri, $qstr); $uri->query_form(@qform, q => $q); my $cmd = $curl->for_uri($lei, $uri); $lei->qerr("# $cmd"); - my ($fh, $pid) = popen_rd($cmd, undef, $rdr); - my $reap_curl = PublicInbox::AutoReap->new($pid); - $fh = IO::Uncompress::Gunzip->new($fh, MultiStream => 1); - PublicInbox::MboxReader->mboxrd($fh, \&each_remote_eml, $self, - $lei, $each_smsg); - if (delete($self->{-sto_imported})) { - my $wait = $self->{import_sto}->wq_do('done'); - } - $reap_curl->join; - if ($? == 0) { - # don't update if no results, maybe MTA is down - my $nr = $lei->{-nr_remote_eml}; + $rdr->{2} //= popen_wr(@lbf_tee) if @lbf_tee; + my $fh = popen_rd($cmd, undef, $rdr); + $fh = IO::Uncompress::Gunzip->new($fh, + MultiStream => 1, AutoClose => 1); + eval { + PublicInbox::MboxReader->mboxrd($fh, \&each_remote_eml, + $self, $lei, $each_smsg); + }; + my ($exc, $code) = ($@, $?); + $lei->sto_barrier_request if delete($self->{-sto_imported}); + die "E: $exc" if $exc && !$code; + my $nr = delete $lei->{-nr_remote_eml} // 0; + if (!$code) { # don't update if no results, maybe MTA is down $lei->{lss}->cfg_set($key, $start) if $key && $nr; mset_progress($lei, $lei->{-current_url}, $nr, $nr); next; } - my $err; - if (-s $cerr) { - seek($cerr, 0, SEEK_SET) // - warn "seek($cmd stderr): $!"; - $err = do { local $/; <$cerr> } // - warn "read($cmd stderr): $!"; - truncate($cerr, 0) // warn "truncate($cmd stderr): $!"; - } - $err //= ''; - next if (($? >> 8) == 22 && $err =~ /\b404\b/); + delete($rdr->{2})->close if @lbf_tee; + seek($cerr, 0, SEEK_SET); + read($cerr, my $err, -s $cerr); + truncate($cerr, 0); + next if (($code >> 8) == 22 && $err =~ /\b404\b/); $uri->query_form(q => $qstr); - $lei->child_error($?, "E: <$uri> $err"); + $lei->child_error($code, "E: <$uri> `$cmd` failed"); } undef $each_smsg; $lei->{ovv}->ovv_atexit_child($lei); @@ -407,9 +386,8 @@ sub query_remote_mboxrd { sub git { $_[0]->{git} // die 'BUG: git uninitialized' } -sub xsearch_done_wait { # dwaitpid callback - my ($arg, $pid) = @_; - my ($wq, $lei) = @$arg; +sub xsearch_done_wait { # awaitpid cb + my ($pid, $wq, $lei) = @_; return if !$?; my $s = $? & 127; return $lei->child_error($?) if $s == 13 || $s == 15; @@ -418,75 +396,60 @@ sub xsearch_done_wait { # dwaitpid callback sub query_done { # EOF callback for main daemon my ($lei) = @_; - local $PublicInbox::LEI::current_lei = $lei; - eval { - my $l2m = delete $lei->{l2m}; - delete $lei->{lxs}; - ($lei->{opt}->{'mail-sync'} && !$lei->{sto}) and - warn "BUG: {sto} missing with --mail-sync"; - $lei->sto_done_request if $lei->{sto}; - if (my $v2w = delete $lei->{v2w}) { - my $wait = $v2w->wq_do('done'); # may die - $v2w->wq_close; - } - $lei->{ovv}->ovv_end($lei); - if ($l2m) { # close() calls LeiToMail reap_compress - if (my $out = delete $lei->{old_1}) { - if (my $mbout = $lei->{1}) { - close($mbout) or die <<""; -Error closing $lei->{ovv}->{dst}: \$!=$! \$?=$? - - } - $lei->{1} = $out; - } - if ($l2m->lock_free) { - $l2m->poke_dst; - $lei->poke_mua; - } else { # mbox users - delete $l2m->{mbl}; # drop dotlock - } + my $l2m = delete $lei->{l2m}; + delete $lei->{lxs}; + ($lei->{opt}->{'mail-sync'} && !$lei->{sto}) and + warn "BUG: {sto} missing with --mail-sync"; + $lei->sto_barrier_request; + $lei->{ovv}->ovv_end($lei); + if ($l2m) { # close() calls LeiToMail reap_compress + $l2m->finish_output($lei); + if ($l2m->lock_free) { + $l2m->poke_dst; + $lei->poke_mua; + } else { # mbox users + delete $l2m->{mbl}; # drop dotlock } - if ($lei->{-progress}) { - my $tot = $lei->{-mset_total} // 0; - my $nr_w = $lei->{-nr_write} // 0; - my $d = ($lei->{-nr_seen} // 0) - $nr_w; - my $x = "$tot matches"; - $x .= ", $d duplicates" if $d; - if ($l2m) { - my $m = "# $nr_w written to " . - "$lei->{ovv}->{dst} ($x)"; - $nr_w ? $lei->qfin($m) : $lei->qerr($m); - } else { - $lei->qerr("# $x"); - } + } + my $nr_w = delete($lei->{-nr_write}) // 0; + my $nr_dup = (delete($lei->{-nr_seen}) // 0) - $nr_w; + if ($lei->{-progress}) { + my $tot = $lei->{-mset_total} // 0; + my $x = "$tot matches"; + $x .= ", $nr_dup duplicates" if $nr_dup; + if ($l2m) { + my $m = "# $nr_w written to " . + "$lei->{ovv}->{dst} ($x)"; + $nr_w ? $lei->qfin($m) : $lei->qerr($m); + } else { + $lei->qerr("# $x"); } - $lei->start_mua if $l2m && !$l2m->lock_free; - $lei->dclose; - }; - $lei->fail($@) if $@; + } + $lei->start_mua if $l2m && !$l2m->lock_free; + $lei->dclose; } -sub do_post_augment { +sub post_augment_done { # via on_destroy in top-level lei-daemon my ($lei) = @_; - local $PublicInbox::LEI::current_lei = $lei; - my $l2m = $lei->{l2m} or return; # client disconnected - eval { - $lei->fchdir; - $l2m->post_augment($lei); - }; - my $err = $@; + my $err = delete $lei->{post_augment_err}; if ($err) { if (my $lxs = delete $lei->{lxs}) { - $lxs->wq_kill('-TERM'); + $lxs->wq_kill(-POSIX::SIGTERM()); $lxs->wq_close; } $lei->fail("$err"); } if (!$err && delete $lei->{early_mua}) { # non-augment case - eval { $lei->start_mua }; + eval { $lei->start_mua }; # may trigger wait_startq $lei->fail($@) if $@; } - close(delete $lei->{au_done}); # triggers wait_startq in lei_xsearch + close(delete $lei->{au_done}); # trigger wait_startq if start_mua didn't +} + +sub do_post_augment { + my ($lei) = @_; + my $l2m = $lei->{l2m} or return; # client disconnected + $l2m->post_augment($lei, on_destroy(\&post_augment_done, $lei)); } sub incr_post_augment { # called whenever an l2m shard finishes augment @@ -509,7 +472,9 @@ sub concurrency { sub start_query ($$) { # always runs in main (lei-daemon) process my ($self, $lei) = @_; local $PublicInbox::LEI::current_lei = $lei; - if ($self->{opt_threads} || ($lei->{l2m} && !$self->{opt_sort})) { + if ($lei->{opt}->{threads} || + defined($lei->{opt}->{'thread-id'}) || + ($lei->{l2m} && !$lei->{opt}->{'sort'})) { for my $ibxish (locals($self)) { $self->wq_io_do('query_one_mset', [], $ibxish); } @@ -531,7 +496,7 @@ sub start_query ($$) { # always runs in main (lei-daemon) process } sub incr_start_query { # called whenever an l2m shard starts do_post_auth - my ($self, $lei) = @_; + my ($lei, $self) = @_; my $l2m = $lei->{l2m}; return if ++$self->{nr_start_query} != $l2m->{-wq_nr_workers}; start_query($self, $lei); @@ -546,17 +511,20 @@ sub ipc_atfork_child { sub do_query { my ($self, $lei) = @_; my $l2m = $lei->{l2m}; + my $qstr = \($lei->{mset_opt}->{qstr}); + chomp $$qstr; + $$qstr =~ s/[ \n\t]+/ /sg; # make URLs and $0 less ugly my $ops = { - 'sigpipe_handler' => [ $lei ], - 'fail_handler' => [ $lei ], - 'do_post_augment' => [ \&do_post_augment, $lei ], - 'incr_post_augment' => [ \&incr_post_augment, $lei ], + sigpipe_handler => [ $lei ], + fail_handler => [ $lei ], + do_post_augment => [ \&do_post_augment, $lei ], + incr_post_augment => [ \&incr_post_augment, $lei ], '' => [ \&query_done, $lei ], - 'mset_progress' => [ \&mset_progress, $lei ], - 'l2m_progress' => [ \&l2m_progress, $lei ], - 'x_it' => [ $lei ], - 'child_error' => [ $lei ], - 'incr_start_query' => [ $self, $lei ], + mset_progress => [ \&mset_progress, $lei ], + incr => [ $lei ], + x_it => [ $lei ], + child_error => [ $lei ], + incr_start_query => [ \&incr_start_query, $lei, $self ], }; $lei->{auth}->op_merge($ops, $l2m, $lei) if $l2m && $lei->{auth}; my $end = $lei->pkt_op_pair; @@ -569,7 +537,6 @@ sub do_query { if ($lei->{opt}->{augment} && delete $lei->{early_mua}) { $lei->start_mua; } - my $F_SETPIPE_SZ = $^O eq 'linux' ? 1031 : undef; if ($l2m->{-wq_nr_workers} > 1 && $l2m->{base_type} =~ /\A(?:maildir|mbox)\z/) { # setup two barriers to coordinate ->has_entries @@ -581,20 +548,19 @@ sub do_query { $l2m->{au_peers} = [ $a_r, $a_w, $b_r, $b_w ]; } $l2m->wq_workers_start('lei2mail', undef, - $lei->oldset, { lei => $lei }); - $l2m->wq_wait_async(\&xsearch_done_wait, $lei); + $lei->oldset, { lei => $lei }, + \&xsearch_done_wait, $lei); pipe($lei->{startq}, $lei->{au_done}) or die "pipe: $!"; fcntl($lei->{startq}, $F_SETPIPE_SZ, 4096) if $F_SETPIPE_SZ; delete $l2m->{au_peers}; + close(delete $l2m->{-wq_s2}); # share wq_s1 with lei_xsearch } $self->wq_workers_start('lei_xsearch', undef, - $lei->oldset, { lei => $lei }); - $self->wq_wait_async(\&xsearch_done_wait, $lei); + $lei->oldset, { lei => $lei }, + \&xsearch_done_wait, $lei); my $op_c = delete $lei->{pkt_op_c}; delete $lei->{pkt_op_p}; @$end = (); - $self->{opt_threads} = $lei->{opt}->{threads}; - $self->{opt_sort} = $lei->{opt}->{'sort'}; $self->{-do_lcat} = !!(delete $lei->{lcat_todo}); if ($l2m) { $l2m->net_merge_all_done($lei) unless $lei->{auth}; @@ -612,34 +578,40 @@ sub add_uri { require IO::Uncompress::Gunzip; require PublicInbox::LeiCurl; push @{$self->{remotes}}, $uri; + $uri; } else { warn "curl missing, ignoring $uri\n"; + undef; } } +# returns URI or PublicInbox::Inbox-like object sub prepare_external { my ($self, $loc, $boost) = @_; # n.b. already ordered by boost if (ref $loc) { # already a URI, or PublicInbox::Inbox-like object return add_uri($self, $loc) if $loc->can('scheme'); + # fall-through on Inbox-like objects } elsif ($loc =~ m!\Ahttps?://!) { require URI; return add_uri($self, URI->new($loc)); - } elsif (-f "$loc/ei.lock") { + } elsif (-f "$loc/ei.lock" && -d "$loc/ALL.git/objects") { require PublicInbox::ExtSearch; die "`\\n' not allowed in `$loc'\n" if index($loc, "\n") >= 0; $loc = PublicInbox::ExtSearch->new($loc); - } elsif (-f "$loc/inbox.lock" || -d "$loc/public-inbox") { + } elsif ((-f "$loc/inbox.lock" && -d "$loc/all.git/objects") || + (-d "$loc/public-inbox" && -d "$loc/objects")) { die "`\\n' not allowed in `$loc'\n" if index($loc, "\n") >= 0; require PublicInbox::Inbox; # v2, v1 $loc = bless { inboxdir => $loc }, 'PublicInbox::Inbox'; } elsif (!-e $loc) { warn "W: $loc gone, perhaps run: lei forget-external $loc\n"; - return; + return undef; } else { warn "W: $loc ignored, unable to determine external type\n"; - return; + return undef; } push @{$self->{locals}}, $loc; + $loc; } sub _lcat_i { # LeiMailSync->each_src iterator callback |