No ext_urls

[public-inbox.git] / lib / PublicInbox / LeiXSearch.pm

diff --git a/lib/PublicInbox/LeiXSearch.pm b/lib/PublicInbox/LeiXSearch.pm
index 2958d3f910b0986baa2bccf138ee185ae35b8c7f..5965274ca211396df49e263ec6ade1508e3ba458 100644 (file)
--- a/lib/PublicInbox/LeiXSearch.pm
+++ b/lib/PublicInbox/LeiXSearch.pm
@@ -1,4 +1,4 @@
-# Copyright (C) 2020-2021 all contributors <meta@public-inbox.org>
+# Copyright (C) all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 
 # Combine any combination of PublicInbox::Search,
@@ -103,13 +103,6 @@ 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 ($$$) {
@@ -163,8 +156,9 @@ sub mset_progress {
 }
 
 sub l2m_progress {
-       my ($lei, $nr) = @_;
-       $lei->{-nr_write} += $nr;
+       my ($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
@@ -285,7 +279,7 @@ sub each_remote_eml { # callback for MboxReader->mboxrd
                my ($res, $kw) = $self->{import_sto}->wq_do('add_eml', $eml);
                if (ref($res) eq ref($smsg)) { # totally new message
                        $smsg = $res;
-                       $self->{-imported} = 1;
+                       $self->{-sto_imported} = 1;
                }
                $smsg->{kw} = $kw; # short-circuit xsmsg_vmd
        }
@@ -376,7 +370,7 @@ sub query_remote_mboxrd {
                $fh = IO::Uncompress::Gunzip->new($fh, MultiStream => 1);
                PublicInbox::MboxReader->mboxrd($fh, \&each_remote_eml, $self,
                                                $lei, $each_smsg);
-               if ($self->{import_sto} && delete($self->{-imported})) {
+               if (delete($self->{-sto_imported})) {
                        my $wait = $self->{import_sto}->wq_do('done');
                }
                $reap_curl->join;
@@ -406,9 +400,8 @@ 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;
@@ -447,13 +440,16 @@ Error closing $lei->{ovv}->{dst}: \$!=$! \$?=$?
                }
                if ($lei->{-progress}) {
                        my $tot = $lei->{-mset_total} // 0;
-                       my $nr = $lei->{-nr_write} // 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 = "# $nr written to " .
-                                       "$lei->{ovv}->{dst} ($tot matches)";
-                               $nr ? $lei->qfin($m) : $lei->qerr($m);
+                               my $m = "# $nr_w written to " .
+                                       "$lei->{ovv}->{dst} ($x)";
+                               $nr_w ? $lei->qfin($m) : $lei->qerr($m);
                        } else {
-                               $lei->qerr("# $tot matches");
+                               $lei->qerr("# $x");
                        }
                }
                $lei->start_mua if $l2m && !$l2m->lock_free;
@@ -577,15 +573,15 @@ 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};
        }
        $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 = ();
@@ -608,34 +604,40 @@ 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