# called by PublicInbox::Search::xdb (usually via ->mset)
sub xdb_shards_flat { @{$_[0]->{shards_flat} // []} }
-sub mitem_kw ($$;$) {
- my ($smsg, $mitem, $flagged) = @_;
- my $kw = xap_terms('K', my $doc = $mitem->get_document);
+sub _mitem_kw { # retry_reopen callback
+ my ($srch, $smsg, $mitem, $flagged) = @_;
+ my $doc = $mitem->get_document;
+ my $kw = xap_terms('K', $doc);
$kw->{flagged} = 1 if $flagged;
+ my $L = xap_terms('L', $doc);
# we keep the empty {kw} array here to prevent expensive work in
# ->xsmsg_vmd, _unbless_smsg will clobber it iff it's empty
$smsg->{kw} = [ sort keys %$kw ];
- my $L = xap_terms('L', $doc);
$smsg->{L} = [ sort keys %$L ] if scalar(keys %$L);
}
+sub mitem_kw ($$$;$) {
+ my ($srch, $smsg, $mitem, $flagged) = @_;
+ $srch->retry_reopen(\&_mitem_kw, $smsg, $mitem, $flagged);
+}
+
# like over->get_art
sub smsg_for {
my ($self, $mitem) = @_;
my $smsg = $ibx->over->get_art($num);
return if $smsg->{bytes} == 0; # external message
if ($ibx->can('msg_keywords')) {
- mitem_kw($smsg, $mitem);
+ mitem_kw($self, $smsg, $mitem);
}
$smsg;
}
pkt_do($lei->{pkt_op_p}, 'mset_progress', @_);
} else { # single lei-daemon consumer
my ($desc, $mset_size, $mset_total_est) = @_;
- $lei->{-mset_total} += $mset_size;
+ $lei->{-mset_total} += $mset_size if $mset_total_est ne '?';
$lei->qerr("# $desc $mset_size/$mset_total_est");
}
}
+sub l2m_progress {
+ my ($lei, $nr) = @_;
+ $lei->{-nr_write} += $nr;
+}
+
sub query_one_mset { # for --threads and l2m w/o sort
my ($self, $ibxish) = @_;
local $0 = "$0 query_one_mset";
my $lei = $self->{lei};
my ($srch, $over) = ($ibxish->search, $ibxish->over);
- my $desc = $ibxish->{inboxdir} // $ibxish->{topdir};
- return warn("$desc not indexed by Xapian\n") unless ($srch && $over);
- my $mo = { %{$lei->{mset_opt}} };
+ my $dir = $ibxish->{inboxdir} // $ibxish->{topdir};
+ return warn("$dir not indexed by Xapian\n") unless ($srch && $over);
+ my $mo = { %{$lei->{mset_opt}} }; # copy
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 $lss = $lei->{dedupe};
+ $lss = undef unless $lss && $lss->can('cfg_set'); # saved search
+ my $maxk = "external.$dir.maxuid";
+ my $stop_at = $lss ? $lss->{-cfg}->{$maxk} : undef;
+ if (defined $stop_at) {
+ die "$maxk=$stop_at has multiple values" if ref $stop_at;
+ my @e;
+ local $SIG{__WARN__} = sub { push @e, @_ };
+ $stop_at += 0;
+ return warn("$maxk=$stop_at: @e") if @e;
+ }
+ my $first_ids;
do {
$mset = $srch->mset($mo->{qstr}, $mo);
- mset_progress($lei, $desc, $mset->size,
+ mset_progress($lei, $dir, $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)) {
my $mitem = delete $n2item{$n};
next if $smsg->{bytes} == 0;
if ($mitem && $can_kw) {
- mitem_kw($smsg, $mitem, $fl);
+ mitem_kw($srch, $smsg, $mitem,
+ $fl);
} elsif ($mitem && $fl) {
# call ->xsmsg_vmd, later
$smsg->{lei_q_tt_flagged} = 1;
@{$ctx->{xids}} = ();
}
} else {
+ $first_ids = $ids;
my @items = $mset->items;
for my $n (@$ids) {
my $mitem = $items[$i++];
my $smsg = $over->get_art($n) or next;
next if $smsg->{bytes} == 0;
- mitem_kw($smsg, $mitem, $fl) if $can_kw;
+ mitem_kw($srch, $smsg, $mitem, $fl) if $can_kw;
$each_smsg->($smsg, $mitem);
}
}
} while (_mset_more($mset, $mo));
+ if ($lss && scalar(@$first_ids)) {
+ undef $stop_at;
+ my $max = $first_ids->[0];
+ $lss->cfg_set($maxk, $max);
+ undef $lss;
+ }
undef $each_smsg; # may commit
$lei->{ovv}->ovv_atexit_child($lei);
}
do {
$mset = $self->mset($mo->{qstr}, $mo);
mset_progress($lei, 'xsearch', $mset->size,
- $mset->size, $mset->get_matches_estimated);
+ $mset->get_matches_estimated);
wait_startq($lei); # wait for keyword updates
for my $mitem ($mset->items) {
my $smsg = smsg_for($self, $mitem) or next;
if (my $lxs = delete $lei->{lxs}) {
$lxs->wq_wait_old(\&xsearch_done_wait, $lei);
}
+ if ($lei->{opt}->{'mail-sync'} && !$lei->{sto}) {
+ warn "BUG: {sto} missing with --mail-sync";
+ }
my $wait = $lei->{sto} ? $lei->{sto}->ipc_do('done') : undef;
$lei->{ovv}->ovv_end($lei);
+ my $start_mua;
if ($l2m) { # close() calls LeiToMail reap_compress
if (my $out = delete $lei->{old_1}) {
if (my $mbout = $lei->{1}) {
$lei->poke_mua;
} else { # mbox users
delete $l2m->{mbl}; # drop dotlock
- $lei->start_mua;
+ $start_mua = 1;
}
}
- $lei->{-progress} and
- $lei->err('# ', $lei->{-mset_total} // 0, " matches");
+ if ($lei->{-progress}) {
+ my $tot = $lei->{-mset_total} // 0;
+ my $nr = $lei->{-nr_write} // 0;
+ $lei->qerr($l2m ?
+ "# $nr written to $lei->{ovv}->{dst} ($tot matches)" :
+ "# $tot matches");
+ }
+ $lei->start_mua if $start_mua;
$lei->dclose;
}
sub do_post_augment {
my ($lei) = @_;
my $l2m = $lei->{l2m} or return; # client disconnected
+ $lei->fchdir or return;
my $err;
eval { $l2m->post_augment($lei) };
$err = $@;
sub ipc_atfork_child {
my ($self) = @_;
$self->{lei}->_lei_atfork_child;
- $SIG{__WARN__} = PublicInbox::Eml::warn_ignore_cb();
$self->SUPER::ipc_atfork_child;
}
'+' => [ \&incr_post_augment, $lei ],
'' => [ \&query_done, $lei ],
'mset_progress' => [ \&mset_progress, $lei ],
+ 'l2m_progress' => [ \&l2m_progress, $lei ],
'x_it' => [ $lei->can('x_it'), $lei ],
'child_error' => [ $lei->can('child_error'), $lei ],
'incr_start_query' => [ \&incr_start_query, $self, $l2m ],
if ($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 &&
+ $l2m->{base_type} =~ /\A(?:maildir|mbox)\z/) {
+ # setup two barriers to coordinate dedupe_nr
+ # between l2m workers
+ pipe(my ($a_r, $a_w)) or die "pipe: $!";
+ fcntl($a_r, $F_SETPIPE_SZ, 4096) if $F_SETPIPE_SZ;
+ pipe(my ($b_r, $b_w)) or die "pipe: $!";
+ fcntl($b_r, $F_SETPIPE_SZ, 4096) if $F_SETPIPE_SZ;
+ $l2m->{au_peers} = [ $a_r, $a_w, $b_r, $b_w ];
+ }
$l2m->wq_workers_start('lei2mail', undef,
$lei->oldset, { lei => $lei });
pipe($lei->{startq}, $lei->{au_done}) or die "pipe: $!";
- # 1031: F_SETPIPE_SZ
- fcntl($lei->{startq}, 1031, 4096) if $^O eq 'linux';
+ fcntl($lei->{startq}, $F_SETPIPE_SZ, 4096) if $F_SETPIPE_SZ;
+ delete $l2m->{au_peers};
}
$self->wq_workers_start('lei_xsearch', undef,
$lei->oldset, { lei => $lei });
$self->{opt_threads} = $lei->{opt}->{threads};
$self->{opt_sort} = $lei->{opt}->{'sort'};
if ($l2m) {
- $l2m->net_merge_complete unless $lei->{auth};
+ $l2m->net_merge_all_done unless $lei->{auth};
} else {
start_query($self);
}